Data Entry Secrets

Free Data
Entry Secrets:
Naked Truth Exposed!
Free Data
Entry Secrets:
Naked Truth Exposed!
Compiled by Michael Bar
MB-Imaging, Delémont Switzerland

Providing authentic secrets that will vastly improve the performance of
your business, this free eBook exposes the naked truth by describing the basic
concepts of data entry market, features of data entry software developed by
the world’s leading manufacturers, and the issue of accuracy provided by
different data entry systems.
You have permission to publish this eBook electronically, in print, or on
your website, free of charge, as long as the following resource box is included
(with the URL linked to our site, when published on the WEB):
Compiled by Michael Bar
MB-Imaging, Switzerland
http://www.mb-imaging.com
If you find its content valuable, please forward the eBook to your industry
colleagues. We would really appreciate it, and so will they!
Also your suggestions and comments in regards to design and content of
this eBook will be highly appreciated.
There is nothing more important to any organization than the accurate
input of data. Today it seems that more money and effort is spent to verify
that data – then the entire process of entering it, is in the first place. We
have clearly identified the problem, so why is the solution so hard to
achieve?
Generally it can be a simple answer: a lack of investment in the proper data
entry tools.
This eBook from MB-imaging outlines many of the issues data entry
departments face today, and tries to address solutions for the reader.
I believe it is a couple of issues: find the experts, invest in the tools, and
use the tools they have developed from their experience.
This book will assist you in that task.
Kevin M. Keyes
Director of International Operations
PHOENIX SOFTWARE INTERNATIONAL
A grateful acknowledgement to my wife Tanja for her
indulgence and goodwill during the writing of this eBook.
MB-Imaging Offers Power/Precision Data Entry Systems
How Important is Accuracy?
Quick Launch The amount of hand-whiting paper documents that are a day-to-day part
Data Entry of your job is on the increase: invoices, checks, bills, orders, applications.
The list is endless. And each document must be reviewed, approved,
Systems
sorted, processed, and ultimately, filed and stored.
Free Many document imaging systems are designed to do most of the indexing
Downloads and data entry by using ICR (or OCR) engines for automatic character
recognition. Only in well controlled circumstances you can benefit from
this technology. However, in the real world, recognition rates seldom
The time-
time- approach the advertised rates:
proven even as low as 1% error rate on a printed page with 3,000 characters
methods of means there is an average of 30 errors on each page!
Unrealized expectations and undetected errors can cause big problems.
key-
key-entering
data result Take a look at the methods your organization uses for data entry:
in 99.99% How important is the impact of errors:
reprocessing costs, unhappy customers, lost business, etc?
Accuracy!
How much it costs your organization to detect and correct errors?
You know today’s ICR/OCR systems have unsatisfactory error
rates for most real-world hand-writing documents.
You know, there must be a better way.
There is!
Our Power/Precision Data Entry Systems

MB-Imaging offers Viking Data Entry Systems – the most powerful today software systems
allowing rapid and accurate keying data from both paper and images:
VDE is a Viking Data Entry System (VDE™) that since 1980 has been the standard for
multi-platform keyboard data entry from paper documents.
VDE allows to enter data up to nine times as fast and with twice the accuracy of
databases, spreadsheets or indexing modules that come with most imaging systems.
ImagEntry is the power software for keying data from images of forms (structured
documents), designed to be the key entry component of high performance document
imaging systems. ImagEntry supplements ICR/OCR/OMR data capture systems with
extremely productive and effective methods to key unreadable characters.
VDE+Images combines the best features of Viking’s VDE and ImagEntry data entry
software, allowing you to key from unstructured documents (which are not amenable to
the successful use of OCR/ICR/OMR recognition engines because the zones of interest are
not definable).
Everything on Your Fingertips

Getting more work done with fewer resources is key to survival these days.
Our Power/Precision Data Entry Systems, perfect for integration into any document imaging
applications, keep vast amounts of data at your fingertips, allowing you to index images and
key-enter data: rapidly, accurately and affordably!
So, why wait? Visit www.mb-imaging.com and find out How You Could Benefit From Our
Power/Precision Data Entry Systems!
Free Data Entry Secrets: Naked Truth Exposed! www.mb-imaging.com
Table of Contents
Why you need Professional Data Entry Software 1

The Importance of Power/Precision Data Entry 5
Captiva Software Corporation 20
FormWare Family
Microsystems Technology 22
OCR for Forms
AnyForm Component
Phoenix Software International 26
Entrypoint Family
FALCON Family
Viking Software Solutions 36
ImagEntry
VDE
VDE+Images
Frequently Asked Questions
Evaluating Data Entry Systems 45
Conversions: What are the Issues for Data Entry Systems? 49
Keying from Images when your Documents are not Forms 53
Top 10 ways to lose money doing Data Entry 55
Related Links 60
Why You Need Professional

Data Entry Software
Microsoft Access, Word, Paradox these are among the software packages many
companies use to enter information into their computers. For small amounts of
data these systems work just fine, as will the standard data entry module that
came with your imaging system, or your own homegrown database front end
written in Visual Basic.
Once you reach a certain volume, however, it makes sense to take advantage of
professional quality data entry software.
Data entry software is designed specifically to speed data entry. Windows-based
software, on the other hand, is designed to be intuitive and easy to use. The
average operator entering data with MS Access, for example, achieves only
about 1,500 keystrokes per hour. Why so slow? Think about how much time is
lost whenever you take your hand off the keyboard to use the mouse.
Professional data entry software uses special technology and dozens of proven
techniques to make keying faster. As a result, professional data entry operators
average keying speeds of 11,600 keystrokes per hour. That’s a 9 to 1
productivity improvement.
Obviously there is a time when you need specialized data entry software. But
how do you know when to move up?
Here are some questions to ask yourself…
How much data are you entering?

There are lots of ways of measuring how much data you enter, such as the
number of records or the number of keystrokes used. But the most relevant
statistic is how much time it takes to key in the data. This is because time is the
measure that translates most directly into how much it costs to enter the data and
allows you to calculate how quickly data entry software will pay for itself.
Let’s walk through that calculation…
Professional data entry software tends to run about $1,000 per user. If we assume
that your fully loaded labor costs are about $20 per hour, we divide $20 into
$1,000 to discover that you must reduce your data entry time by 50 hours per
operator to save enough money to pay for the software ($1,000 / $20/hour = 50
hours).
1
We can reasonably estimate that you will get a 30% productivity increase from
data entry software. So we divide 30% into 50 hours to find that you will need
167 hours of keying to recoup the software cost (50 hours / 30% = 167 hours).
If you have people spending an hour a day keying in data and correcting keying
errors, a data entry system will pay for itself in less than a year (167 hours / 20
hours a month = 8.4 months). If your employees are keying 2.5 hours a day,
you’ll reach break-even sooner – in about three months (167 hours / 50 hours a
month = 3.3 months).
How much do you pay your employees?

For most companies, labor costs overwhelm the cost of data entry software. This
is particularly true if you are in a high cost labor market or have highly paid
employees who spend part of their day entering data.
For example, if you are paying the people who enter data $25,000 to $30,000 a
year (including benefits and taxes), and you achieve only a 5% gain in
productivity you can save $1,500 per year. This is enough to pay for professional
data entry software. Most companies see much higher productivity gains than
5%, so the software becomes easier to justify. And these savings continue year
after year.
Don’t forget to factor in the labor costs of setting up and maintaining data entry
applications. If you are using highly paid programmers, you may find that you
can free them up for other projects when you get data entry software. Some
systems are so easy to use that operators can set up new jobs and maintain the
system.
How important is accuracy?

The more important accuracy is, the more seriously you should look at
professional data entry software. In addition to increased productivity, one of the
major benefits of data entry software is error reduction.
If your data is text that will be read but not processed further, then errors may not
be important. Readers can get the meaning from the context. However, if you
will be using the data in other applications, the old “Garbage in, garbage out”
rule applies. You need to consider the impact of errors: reprocessing costs,
unhappy customers, lost business, etc.
You also need to look at how much it costs your organization to correct errors.
Usually, it is almost free if you catch an error at the time the finger is on the key
and the eye is on the information. It may cost $20-$100 to correct it after it has
been keyed if you have to go find the original documents and create updates to
2
the data. It is even more costly when the bad data has been propagated
throughout the system and has to be corrected in more than one place. And don’t
forget those intangible costs associated with customer goodwill and lost
business.
How does professional data entry software increase data accuracy? Most systems
incorporate error detection at three levels: character level, field level, and record
level. They also provide double key verify capability.
Character Level Validations

Character sieves disallow certain characters, e.g., no letters are permitted in
numeric fields. Good systems will provide a variety of character sieves. These
must be instantaneous so they do not slow down the user.
Field Level Validations

Extensive field edits help you catch errors that can be detected as soon as the
field is completed. Some examples are range checks, date validations, database
lookups and multi-field calculations. Professional quality systems allow you to
do most, if not all, of your field edits with out having to write program code.
Record Level Validations

Some errors cannot be detected until all of the data for a transaction has been
entered. These validations must be essentially instantaneous or else the user
cannot reach his or her potential productive speed. The old mainframe data entry
where you hit ENTER and waited and waited and waited… is unacceptable.
Financial data in particular needs “balancing” validations.
Double Key Verify

The time-proven method to ensure accuracy of key-entered data is to key it
twice, preferably by different people, and compare the results. Key verified data
is 99.99% accurate, on average. Professional data entry systems have quick and
easy methods for the Verify Operator to make corrections. There is no need to
double key verify fields that can be validated by other methods, so this feature
must be optional by field.
Summary
Take a look at the methods your organization uses for data entry. Run your own
labor costs through the calculation given above, and factor in the importance of
data accuracy.
3
You should be able to tell if your company can benefit from acquiring
professional data entry software. Or turn to a data entry software vendor for
assistance with the analysis. You can also take advantage of their experience to
obtain additional insights into maximizing data entry productivity gains for your
organization.
4
The Importance of Power/Precision

Data Entry to Document Imaging
Introduction
Cost of Quality
ICR/OCR Accuracy Issues
Miscellaneous: Other Factors that Influence Recognition Rates
Other Issues
Conclusion
Introduction
One of the often overlooked and underestimated elements of a Document

Imaging project is the automation and human cost of lifting data from the
images. Systems engineers frequently respond to data entry requirements with
automation-only solutions that fail to recognize the accuracy demands of most
applications. We have all heard horror stories about lost images due to improper
indexing and major errors due to inaccurate invoices, claim forms and etc
The off-the-shelf key entry component of most document imaging systems is
seldom up to the job. They lack the attention to data entry cultures and
automation costs that are critical in high volume, high accuracy data entry
operations. Ongoing labor costs for data entry are often the largest element of the
operating costs. Unnecessary operating costs are the result of failure to provide
state of the art data entry methods and techniques. Controlling these costs can
mean the difference between economic success and failure of imaging projects.
Most data entry modules for image systems were designed by Windows
programmers and are excellent for casual use. Nevertheless, they are inefficient
for high volume, high speed data entry. For decades the national average for
keyboard data entry has been about 12,000 keystrokes per hour. Anything less is
unacceptable to organizations concerned about costs.
The professional data entry people in your organization can offer valuable
insights into these issues and their importance. Be sure to involve them in the
evaluation process.
5
Many document imaging systems are designed to do most of the indexing and
data entry by a Akeyless@ method using ICR (or OCR) engines for automatic
character recognition. In well controlled circumstances this technology will
reduce the key-entry labor dramatically. However, in the real world, using
uncontrolled documents, recognition rates seldom approach the advertised rates.
Unrealized expectations and undetected errors can cause big problems. The
remainder of this paper will address the issues and explain the importance of data
entry methods and techniques that will be required by successful document
imaging systems. The discussion is divided into five categories:
Importance of Data Accuracy
Operating Costs
Application Development Costs
Other Issues
Cost of Quality
Errors in data cause a variety of problems and raise the costs in several areas.
The cost to recognize and detect errors is not trivial. Further costs are incurred to
correct the data errors. The largest cost components are the hidden costs that
affect other departments.
Error Detection
Detecting data errors in programs often takes as much, or more, of the analysis
and programming efforts than the main logic. The earlier an error is detected, the
cheaper it is to correct it. Fifty years ago it was common to Akey verify@ punch
cards. This technique is still one of the best methods for detecting errors. Re-key
verifying selected data items, combined with programs that look for invalid data,
can detect most data errors.
Error Correction
Correcting errors programmatically is the cheapest way. Doing it with a single
keystroke when the data is initially keyed is the cheapest manual method.
Conversely, it costs hundreds, or thousands, of times more to create update
transactions to fix errors that remain undetected until later in the process.
6
Hidden Costs
The hidden cost of errors is much higher. For example, customer service
problems increase proportionately to the number of mistakes that result in billing
errors or shipping the wrong product. You know the many hidden costs in your
business.
A whole specialty knowledge area has developed around how to design and tune
character recognition applications. The state of the art is constantly improving,
but it still requires very scarce technical resources to be successful. It is beyond
the scope of this paper to discuss ICR and OCR technology. However, we will
discuss some of the error detection and correction issues.
Many ICR systems claim 95% to 99% data accuracy. This sounds great on the
surface, but in fact may be unacceptably low. When one understands that this
rate is on a “per character” basis and that ICR engines reject characters that don’t
meet image quality specifications, these advertised rates become increasingly
unappealing. Only 95% accuracy means 5% errors. A typical data input
transaction consists of more than 60 characters. How many applications that
have an average of three errors on each transaction would be considered
successful?
Error Types
There are two classes of recognition errors, unrecognized characters (called
Arejects@) and erroneously recognized characters, or Asubstitution@ errors.
Rejects need to be corrected. Substitutions must first be detected and then
corrected.
ICR engines can achieve very high recognition rates when the documents are
properly designed, printed and controlled. Exploitation of corporate data and use
of “fuzzy logic” expert systems can further boost ICR results. Nonetheless, a 1%
error rate on a printed page with 3,000 characters means there is an average of 30
errors on each page. This would be unacceptable for a typist, but may be
adequate for information that will only be read occasionally and if the human
reader or text classification engines can deduce the correct information from the
context.
7
However, today’s ICR/OCR systems have unsatisfactory error rates for most
real-world documents. It is a continuously developing technology, but it is
projected to improve only incrementally for the next few years. Furthermore,
data to be processed by a computer must be very accurate for most applications.
The true measure of the value of ICR is the cost of doing it plus the cost of error
correction. This is highly application dependent. Many ICR projects are highly
successful, but others cost more than simply keying the data.
Reject Correction Costs

On the surface, reject reentry seems simple and straight forward. An image of the
rejected character is presented to the keyer who depresses the correct character
key and the program moves to the next reject character. Reject reentry modules
must address three problems:
missing characters
extraneous characters
key entry speed
Missing Characters
Sometimes the recognition engine cannot separate adjacent characters and treats
them as a single unreadable character. In this case the operator must be able to
insert extra characters. The operator must take extra time to diagnose the
problem, which affects the throughput. This process must be ergonomically
efficient to achieve maximum keying speeds.
Extraneous Characters
Smudges and other extraneous marks on the paper may appear as an additional
unrecognized character. The operator must decide what the problem is and delete
the extraneous character. Again, the ergonomics of the program design strongly
influences effective keying speed.
Key Entry Speed

Data is usually keyed from paper documents at average rates between 10,000 and
15,000 keystrokes per hour. Keying from images should be slightly faster since
the keyers do not have to remove their hands from the keyboard to turn the pages
of the source documents.
With some systems, reject reentry keying rates may be only 5 to 10% as fast as
full keying rates.
8
One reason for the slower keystroke rate is that presenting keyers with only a
single character at a time prevents them from building any rhythm. Smooth
keying rhythm is essential to fast keying.
Costs
Two Cost Factors

We see that two factors influence the cost to repair rejects;
Accuracy of the recognition engine, and
Reject reentry speed.
The facts show that if the recognition engine is only 90% accurate and the reject
reentry rate is only 10% as fast as full keying, it is just as cheap to key the data.
On the other hand, increasing the recognition to 95% will make it very
worthwhile. Increasing the reject reentry rate also leads to substantial cost
reductions.
Since reject characters often occur together, the ability to simply re-key the
entire field, or even the entire record, greatly speeds up reject reentry rates. This
is because the operator can often key the entire field faster than repairing several
rejects. This capability is very valuable for forms and documents that are subject
to being damaged, marked, stained, or wrinkled.
Reject repair programs for high volume applications must be well designed to be
cost effective. The value of recognition engines depends on many factors and it
also varies from one application to another.
Substitution Error Costs

We must first consider the error detection component. Correcting rejects is easier
than substitutions because the recognition engine finds the errors for us. With
substitution errors we have to find the errors that fooled the recognition engine.
Some errors can be detected programmatically with spelling checkers, data base
lookups, deduction from other data, balance calculations and semantic analysis.
Programmatic error detection is highly application dependent.
Manual Detection Methods

There are also two manual error detection methods. Proofreading (or sight
verification) is the most common method. It is not especially accurate because
9
the mind has a way of fooling our eyes. Nevertheless, it finds many errors.
Surprisingly, continuous proofreading is not as fast as key entry.
Some reject reentry systems provide for selected proofreading while the rejects
are being corrected. This can be a valuable feature.
Re-key verify is the time proven method of manual error detection. Over decades
of use it has been proven to be about 99.9% accurate. The cost is similar to the
cost to key the data. However, usually not all of the data has to be verified which
saves labor. Data elements that can be programmatically validated, or whose
accuracy is not important, do not need to be key verified. Good data entry
programs have this capability. Verifying only sample portions of the data is a
statistical method used to detect problems with equipment and personnel.
Correcting Substitution Errors

Once the error has been detected, the issue of how it is to be corrected must be
addressed. Some systems allow the error detection program to Amark@
substituted characters as a rejected characters and send the data record and its
image back through the reject reentry process. Re-key verify programs should
provide an efficient method for immediate error correction. Some are more
efficient than others.
Miscellaneous: Other Factors that Influence Recognition Rates
Recognition accuracy depends on many factors. Good quality documents with

well delineated text is the first requirement. Paper color and type also affect the
quality of the scanning and resulting image.
Good text delineation implies that cursive handwriting is beyond the scope of
today’s technology except for very highly specialized cases where the system
can Alearn@ the unique characteristics of the author's handwriting.
Document Recognition
Document recognition is clearly a prerequisite for formatted forms. If the
document type is unknown then the recognition engine will not know where to
look for the data. This may mean that not only does the type of form need to be
recognized, but also the particular version of it. In these instances, character
recognition doesn’t help much with “content recognition”. When forms
recognition is difficult, it may be cheaper just to key enter all the data. Or, plan
10
to do manual entry in the beginning until satisfactory document recognition

technology is in place, the forms are redesigned, and the old forms have been
removed from circulation. The point is, plan your image data entry for reality.
Character recognition technology is changing rapidly. Proprietary systems that
are locked into a single recognition engine may result in less than optimal
performance today. They risk falling behind the technology curve in the future.
No one knows what will be the hottest engine next year.
Recognition Technology
Some engines are much better on some kinds of applications than others. If you
have a variety of documents it may be desirable to consider multiple engines.
Operating Costs
Operating costs go on for the lifetime of the image project. Typically they will be
orders of magnitude higher than the acquisition and implementation costs. Some
operating cost elements associated with data entry have been previously
mentioned. The four principal issues are:
Maintenance Costs
Data Accuracy
Keying Speed
Controls and Auditing
Maintenance Costs
What are the annual maintenance fees for licensed software?
What will it cost to maintain the hardware? Systems that depend on highly
specialized components not only cost more to acquire, but their maintenance
costs will be much higher than standard components. For example, high
performance workstations and monitors usually must be maintained by the
manufacturer, while common PC’s are inexpensive to maintain, replace and
upgrade. Power consumption and other environmental factors must be
considered.
How much will it cost to maintain and enhance products developed in-house?
The tremendous growth of the packaged software industry speaks to the fact that
it is usually much cheaper to maintain licensed software than to maintain
systems developed in-house. This is especially true for dynamic new
technologies like imaging. You expect your software vendor to keep up with the
11
advances in technology and improved methodology. Are you willing to make a

similar commitment of in-house resources? If not, you will soon fall behind and
you will not achieve the cost savings to which you are entitled and which
management expects.
Data Accuracy
The importance of this topic was previously discussed. Now we will discuss
some techniques and methods the data entry software should offer to allow you
to maximize the accuracy of your data. These are features that will be found in a
good key-from-image program.
Validation Features
Character Sieves. The first line of defense against keying errors is to eliminate
invalid keystrokes as they occur. For example, the system should not allow alpha
characters in numeric fields and provide for even more sophisticated single
character validations.
Field Edits. As soon as a field is entered, it needs to be edited. You need a wide
variety of edits, ranging from simple numeric range checks to database lookups
and computed values. The common functions should be provided by the system
and there should also be a mechanism for easily adding custom edits that are
unique to the application.
Field edits should have multiple levels of capability. For example, the first level
date validation is to insure the month is in the range from one to twelve. A
deeper level detects future dates and unreasonably old dates.
High Speed Table Lookups. Comparing field values with database tables of
acceptable values and other information is one of the most important types of
field edits. Table lookup with substitution is a valuable Akeystroke saver@ that
improves productivity. However, this process must be extremely rapid and
efficient and never delay the keyer. Table lookups that disrupt keying rhythm
may be counterproductive. This requires very good programming techniques and
well-designed systems to be effective.
Computed Values. To achieve maximum productivity and accuracy the system
must provide for flexible and fast field edits to compute values based on one or
more data fields and database values.
Check Digits. Many fields, such as account numbers, have a self-checking digit
in the number that can be used to detect errors in the value. Dozens of check
12
digit algorithms are in use and the data entry module must be able to handle them
all.
Field Duplication. A common method of keystroke reduction is the capability to
duplicate the values in previous fields and from previous documents. Often the
documents are grouped by location, date, account number and other fields that
can be duplicated. Default values are a special case of duplication. The system
should provide for a flexible and powerful way to duplicate data. The keyer and
the system both should be able to duplicate information from other fields.
Context Sensitive Help. The system should provide on-line help to guide the
keyer in the rules for entering data. Help messages should present information
for the data field being keyed. It should require a minimum of keystrokes to
request and terminate the help messages. Multilevel help messages with
increasing orders of detail are even more useful. It must be easy to modify and
change the help messages because data entry forms are subject to change and the
needs of the keyers may not be well defined initially.
Optional and Required Fields. Some fields must always be entered and the
system must enforce this requirement. Other fields are optional. Some optional
fields are seldom entered so keyers will be more productive if the default is to
automatically skip them. But there must be an efficient way for the keyer to enter
the fields when necessary.
Intelligent Field Skipping. The best systems have the ability to dynamically skip
fields based on the data entered in previous fields. This feature speeds key entry
and reduces the opportunity to make errors.
Re-key Verify
As mentioned previously, the time proven technique for improving data accuracy
is to independently re-key the data and compare the results. Accuracy is much
better when the keyer doing the verify step is not the one who originally entered
the data. To be cost effective, this process must be very efficient. Not every field
needs to be verified, so selective verification is required. The system should
recognize certain conditions for dynamically skipping verify fields to further
reduce the effort.
Usually the verify keyer is more experienced than the original entry keyer and
should be able to correct the data. The correction process must be very efficient
in order to minimize the time to verify data. For example, single character errors
13
should be easily corrected when they are detected. Likewise, an entire field may
have to be replaced.
Re-key verify is also used as a means to measure the results of programmatic
validation techniques and tune them for improved performance.
Keying Speed
There are many factors that influence keying speed. There is no single overriding
factor, but a whole series of little things that couple together to allow the keyer to
reach their maximum potential. Interestingly, studies have shown that the fastest
keyers are also the most accurate. This means that these ergonomic factors are
important even if you do not expect blinding speed from the keyers.
Fully keyed data entry is accomplished at average rates of 12,000 keystrokes per
hour. Superior operators will key at 50% or more above the average rate. They
can only do this if the data entry system is well designed, allows the keyer to
maintain rhythm, and minimizes hand movement. The techniques and methods
for high speed key entry are evidently not widely understood and appreciated. A
new generation of analysts has grown up with a mouse and has no concept of
what is involved in keying at 20,000 keystrokes per hour. Production data entry
is very different from the casual data entry we use with spreadsheets, data bases
and other transaction-oriented applications.
Keying Rhythm
Rhythm is a very important factor in high speed keying. The best operators talk
about the importance of maintaining a smooth rhythm to the keying.
Furthermore, fast keyers actually look ahead of where they are keying. There is a
discernible time lag between the eyes and the fingers.
Keypunch Style Keyboard

The fastest keyers traditionally have used the so-called A029 keypunch@
keyboard layout that has the numeric keys underneath the right hand. This allows
numeric characters to be keyed at the maximum rate without moving the hands
from the home keys. This keyboard has proven to be the fastest for alphanumeric
data. Good data entry systems emulate this capability on a standard keyboard.
Enter Key vs. Tab Key

Data entry applications have used the ENTER key to signal the completion of a
data field. This is a large key under the strong right hand.
14
Windows applications use the TAB key for this purpose. It is a small key located
under the left hand. This use of the TAB key makes it unavailable for Atabbing@
over optional fields.
Removing the hand from the keyboard to manipulate a mouse to perform special
functions will not be nearly as efficient as assigning the special functions needed
by data entry keyers to a function key that can be depressed with a single
keystroke.
Statistics
Statistics should be kept regarding the number of errors and who made them.
This is useful both for evaluating the keyers and to refine the system. Reports
should be available that are organized by operator and by job. It has often been
said that, if you can’t measure it you can’t manage it@.
Traditional data entry systems have measured the keyer's entry rate in keystrokes
per hour. This metric must be carefully considered as it may not be applicable
when entering data from images or when correcting rejects from recognition
engines. However, it can be a good comparative measure for keyers doing the
same work. A better metric may be “error-free records per hour”, or “documents
per hour”. “Documents per hour” is better for evaluating changes to the system.
It is not possible to compare average keystrokes per hour between different
systems. The method of allocating the time to keyers varies from one program to
another, so it is an Aapples and oranges@ problem.
Controls
The data entry supervisor must be able to know what jobs are in the system and
where they are in the process. They must be able to control the work assignments
to keyers. Controls must be flexible because priorities change as deadlines
approach. The system must be auditable to ensure that all the work coming in has
been processed and is not lost. It is desirable to know if it is being accomplished
in a timely fashion.
File/Batch Navigation
The system must provide facilities for the keyer to view the various images and
data records in a batch of work. They should be able to easily go backwards to
review previously entered work.
15
Search Mechanism
A flexible search mechanism is needed to find records and images based on the
contents of the data.
Modify Existing Data

The keyer must have the ability to change data that was previously entered. This
includes both key entered data and data from recognition engines. There should
be safeguards to prevent data from accidentally being changed.
Character Insert/Delete
The keyer must have the ability to insert and delete characters. The most efficient
mode for high speed data entry is typeover mode where characters are replaced
by typing over them. However, when correcting existing data it is desirable to
select character insert mode so that characters can be easily inserted and deleted
from the text. Good systems provide both modes.
Image Display Speed

The system should never prevent the keyer from reaching their maximum keying
rate. Some keyers are capable of speeds in excess of 20,000 keystrokes per hour.
Simple tests, such as holding down a repeat key, can often be used to determine
if the data entry system can display images fast enough.
Image display speeds are often the limiting factor and they should be studied to
determine the effect on overall productivity. Image display speed depends on
many factors, such as:
CPU speed of the workstation
Display monitor
Image resolution
Display parameters: e.g., color, grey, etc.
Decompression algorithms
Application Development Costs

The costs to develop the data entry component of document image systems vary
widely. Some systems provide easy-to-use tools that allow you to quickly and
easily create the applications. Other systems use closed and proprietary
technology that requires extensive training and sometimes you must use their
consultants to set up the applications.
16
Systems that use the same application development procedure for keying and
reject reentry are simpler to use and also have a consistent interface for the
keyers.
Other Issues
There are many other issues that affect the performance of an image data entry
system. Space does not permit a thorough examination of these factors and they
will only be mentioned briefly.
Operating System
The most common operating system for an image data entry platform is MS
Windows. Most of the imaging components are designed for Windows. The
familiar interface makes it easy to learn to use systems and to integrate
components. Although Windows is inherently much slower than other operating
system, today's high speed PCs generally offset this disadvantage. However,
standard Windows dialog boxes are inadequate for professional data entry
operators.
It is crucial that the key-from-image module provide the features and
functionality to enable the data entry operator to key at their maximum potential
speed. Rates far in excess of 12,000 keystrokes per hour are routinely recorded
by professional data entry operators. Be certain that your system will provide this
level of performance.
Interactive terminals connected to a multi-user UNIX system can be used as data
entry workstations. However, they are usually not price competitive.
Block mode terminals attached to mainframe computers are not nearly as
productive as character interactive systems and they cost much more to procure
and operate. Data entry is a highly interactive activity.
DBMS Compatibility
The data entry system must provide some compatibility with the DBMS used by
the rest of the image system. It may only need to be able to import/export data
files.
Proprietary Issues
Systems that are tightly integrated to proprietary hardware and software
components have advantages and disadvantages.
17
On one hand, proprietary systems tend to be the most uniform and there is a
single point of contact. On the other hand, systems composed of Abest-of-
breed@ components are more likely to have the best performance and lowest
cost. As in all things, there are always gray areas in between these two extremes
to consider.
Multiple Image Formats

The trend in the industry is towards TIFF formats, but there are many other
formats in common use. Good data entry modules can handle a variety of
formats interchangeably.
Workflow
The data entry component must provide workflow capability for images and data
that are in the various stages of data entry. It should be mostly automatic with
manual overrides by the supervisors.
The system should also be compatible with the workflow mechanism used by the
rest of the document imaging system. Hooks and handles should be available to
provide the proper interfaces.
Flexibility and Adaptability

The system should be flexible to accommodate growth and change. New
applications will arise and new technologies will be added to the system.
Recognition engine technology is on a steep upward curve. The best engine
today may be superseded by something much better tomorrow. Data entry
components must adapt to rapid change. The system should be able to integrate
with stand-alone data entry components.
Importing Data
A flexible method to import data into the system is clearly needed. Systems that
are locked into single data formats and databases are of limited value. This is
especially true in the case of multiple recognition engines.
Exporting Data
After the data entry function is complete, the data must be exported to the
downstream applications and databases. Portions of the data may be exported in
different formats. A flexible export module is required.
18
Conclusion
Data entry often contributes the largest cost component for document imaging
systems. The costs can be classified in the following three categories:
Data Capture Costs
Error Detection Costs
Error Correction Costs
Hidden Costs
Well-designed systems can minimize these costs, but attention to detail and time-
proven ergonomic techniques are essential. Keyless data entry technology is
improving, but manual data entry and reject reentry is going to be required for
years to come. Organizations will not be well served by glossing over this issue.
Rather, they should be certain that the data capture components of their
document imaging systems are well suited to their requirements. Superior data
entry modules are available as off the shelf products.
19
Captiva Software Corporation
FormWare™ Family
FormWare is designed to serve complex capture needs in higher-volume, multi-

user environments to fill the bills with outstanding scanning, fax capture,
recognition, key entry, and export capabilities. FormWare also excels in its
extensibility – complex business rules and routing can be plugged in to expand
functionality or alter the flow of information. FormWare is commonly found in
environments where a large portion of the documents processed is not “ICR-
friendly”.
FormWare offers a universal capture platform to manage business-critical
information from forms, documents and electronic data. It is used to extract
information from forms of all types, including medical and other insurance
claims, invoices, credit applications, tax returns, order-entry forms, etc.
Desktop is fully expandable, entry-level version of FormWare. It combines

document capture and forms processing in one easy to use package, perfect for
low volume document capture and forms processing. Desktop supports a wide
range of capture applications, from document imaging to forms processing
(OCR/ICR) and data key-entry.
It comes pre-configured with the Desktop Starter Pack. This package includes
five typical customer applications and workflows that can be used as templates
to make it easy to customize applications to any specification. And with image
enhancement, quality control and Key from Image (KFI), Desktop minimizes
the need to re-design most forms.
InvoicePack uses innovative optical character recognition (OCR) and neural

network technologies to extract critical data from invoices and transfer that
information to ERP, accounting and document imaging systems with minimal
human intervention. To ensure the value and integrity of all information
captured, InvoicePack includes a suite of user definable business rules that
guarantee data accuracy and validity and leverage legacy data in the existing
systems.
After an invoice is scanned or received as a fax, the software automatically de-
skews and “cleans” the image in order to minimize unnecessary “noise” that
might interfere with subsequent processes.
20
Next, InvoicePack automatically attempts to identify the invoice type by

referring to a database of previously encountered and known invoices. Assuming
the invoice is confidently identified, an associated template is applied to
optimize OCR results.
If the invoice type can’t confidently be identified, InvoicePack uses OCR
technologies to recognize all data on the invoice and parse through the results to
extract the desired information. Since no OCR technology ever provides a 100%
solution, accounts payable staff then use InvoicePack’s high speed data entry
interface to perform reject repair on erroneous or suspicious characters or
manually key entire fields of data from the invoice image. In addition, user
definable business rules are constantly applied throughout these processes to
ensure the accuracy and validity of all data captured.
InvoicePack comes with a set of pre-defined templates designed to address
many common invoice types. This database automatically grows as new invoice
types are encountered. The software uses neural network technologies to
automatically “learn” those invoices “on-the-fly” and store the resulting
templates. Therefore, the system is constantly becoming more intelligent and
improving its own performance while letting you avoid the cost prohibitive task
of defining a new template for each unique invoice. Users can manually define
new invoice types or fine-tune templates in order to optimize invoice
identification and OCR results.
Free Form Processing allows users to capture and process data from structured
and unstructured forms. The Free Form solution is a cutting-edge technology
that employs fuzzy logic to locate particular data types on unstructured forms
(invoices, EOB’s, purchase orders and more) and extract required information.
Historically, forms processing programs have required predefined templates in
order to process documents. However, with FormWare’s Free Form capabilities,
users can capture data from unstructured forms. This functionality is extremely
valuable since it is common for the location of data within a document to vary
from form to form. Free Form technology can capture information regardless of
its location on a document.
The user is able to define key word searches in a particular zone area of the form
and extract the data from varying positions. These data types may be similar in
content, but are located in different sections of the form as commonly found on
invoices and purchase order forms.
21
Microsystems Technology
OCR for Forms
OCR for Forms reads data from any scanned form and works with most
standard production scanners, including color and bi-tonal (black and white).
OCR for Forms is designed to maximize scanner features that improve image
quality based on logical groupings of forms. If the images are created by another
means (e.g. another imaging system, faxes, etc.), they can be imported into the
OCR for Forms’ environment. In addition to processing standard forms, OCR
for Forms can handle multi-page and duplex forms (e.g., 10-page questionnaires
or surveys), as well as booklets, drop-out forms, color forms and even complex
“formless” forms with ease. OCR for Forms’ first level of image enhancement
is applied to improve image quality. The software is designed to maintain or
improve the images it handles – an important fact, given the importance of high-
quality images in forms processing, as well as storage and retrieval.
Before any other processing begins, an optional “Quality Assure” step can
immediately assess the quality of the scanned or imported images. Each image
can be visually inspected in detail. If the image is too light, dark, noisy, etc., the
scanner settings can be adjusted and you can selectively re-scan to get the best
possible image. For those not-so-perfect forms, you can enhance the image
utilizing “Image Cleanup,” including de-skew, de-speckle, character-smoothing,
de-shading, etc. “Image Cleanup” can be performed on images originally
imported as well as scanned. In addition, you are able to prioritize the batch for
immediate processing, which will place it ahead of all other batches in the queue.
Once a paper document is scanned, imported or faxed into the system, OCR for
Forms detects, identifies and routes the data image to the proper PC station for
processing. OCR for Forms’ automatic form identification eliminates the time-
consuming work of sorting forms. In addition to automatic form identification
based on the topology of the form, you have the ability to supply patch or bar
codes to differentiate batches of forms, or you can define an area on the form for
specific form identification. OCR for Forms can even identify document sets
and multi-page images. When a document contains both forms and attachments,
OCR for Forms will carry all attachments through the system. These
attachments will then be saved with the corresponding batch.
22
When a form cannot be identified, OCR for Forms first automatically rotates
the form 180 degrees to determine if it was scanned upside down. If the form is
still unidentifiable after this process, you still have a number of options available
for handling, from automatic routing to manual exception processing.
OCR for Forms typically has its greatest impact in reducing your processing
costs when it automates work that was previously performed manually. During
this process, the system executes form removal, image enhancement and
character regeneration. These technologies enable you to process your existing
forms, removing distracting clutter (lines, patterns, background noise, etc.),
thereby maximizing the readability of your data and allowing the system to work
more efficiently with greater accuracy. In addition, OCR for Forms now offers a
premier voting capability to improve recognition results and guarantee greater
accuracy for users. OCR for Forms’ recognition engines process machine print,
check boxes, bar codes and hand print, and has language recognition of more
than 20 countries.
Data capture software must be both flexible and precise. Therefore, the system
will interpret these formats according to your specific requirements, assuring
clean characters are used in the recognition process. OCR for Forms excels at
processing even the most complex forms.
Accurate data regeneration and recognition is only one of the ways OCR for
Forms’ technology expedites forms processing. Another way is through Data
Validation, which minimizes the number of characters that need to be reviewed
by an operator during the verification phase. By limiting the amount of manual
intervention that needs to be performed, this technology further speeds up
processing. Through a series of pre- and post-processing rules you define for
your specific business, even questionable or incorrect entries can be validated,
corrected and captured automatically. All this occurs before a single person
intervenes, other than to place the forms in the scanner.
Data verification is a high-speed, heads-up procedure confirming all data entered

is correct and complete for optimal accuracy. During verification, a user visually
reviews and corrects questionable characters and fields that do not conform to
your business rules. The important benefit is instead of manually entering 100%
of the data from paper, you only verify a few questionable entries; the rest is
automatic. In addition, you are able to target a specific user to receive a
particular form for verification.
23
This allows each user to focus on one type of document for consistency and
increased productivity. You also have the option of defining critical data fields
for mandatory verification to ensure the data is accurate. The system was
designed for maximum user productivity, and delivers fast, accurate processing
results with color or black-and-white forms. Hitting just one key allows you to
move from questionable character to questionable character. Since the system
will automatically move through all of the images in the batch; there is no need
to move your hands from the keyboard. In addition, you have the ability to move
around within the images by paging up or down.
Double-Key Verification
Double-Key verification is an optional parameter allowing you to perform
additional verification on those fields that are crucial for accuracy. You can also
define a different user for the second verification, guaranteeing that the same
mistake isn't made twice. It is a time-proven data entry method to key the data
twice in order to help ensure 99.9% data accuracy.
Batch Balancing
OCR for Forms allows you to easily verify the integrity of a batch of forms.
Simply perform a calculation from various fields across the forms and match the
total to a batch separator page or a keyed-in value from the scan operator. In
addition to performing an equation at the batch level, you can perform it for a set
of documents. For example, a multiple-page invoice where you want to verify
that all line items have been accurately extracted and match the total at the
bottom of the invoice. If the equation doesn’t balance, you will be given an
opportunity to make corrections during verification with an easy-to-use “batch
balancing” screen.
The final step is for OCR for Forms to generate an output file and/or an image
file for future reference. OCR for Forms provides flexible tools to generate an
ASCII file in almost any format required, including the option to save scanned
information to a PDF file.
Applications (structured forms):

Time Sheets
Purchase Orders
Inventory / Order Forms
24
AnyForm Component
AnyForm Technology incorporated special features to process the unique

characteristics of an Invoice. Because no template is required, AnyForm uses
pre-defined keywords, word filters and retainers, and the unique topography of
the Invoice to locate the correct data, wherever it is located on the Invoice.
You also have the ability to apply specific business rules to the information
found on the Invoices (e.g. purchase order numbers must be 10 numeric
characters beginning with “2”, or payment terms should be net 30 days, etc.).
Because AnyForm is a component of the OCR for Forms solution, you receive
all the benefits of this combination of technologies.
Applications (unstructured forms):

Invoices / Credit Memos
Bills of Lading
Shipping / Mortgage Documents
25
Phoenix Software International
Entrypoint Family: Entrypoint for Windows

Entrypoint Plus (with special features for clinical trials)
Entrypoint 90 for MS-DOS
Entrypoint
Entrypoint is a complete application development package for creating high-

performance data entry applications in the Windows operating environment.
Entrypoint consists of Application Developer, Data Entry Runtime
(workstation), and Administrator Modules. Entrypoint interfaces with standard
communications products and commercial networks.
Quick, Easy Development

Entrypoint allows users to quickly and easily create a wide variety of data entry
applications with built-in validation, editing, and export options.
Entrypoint contains many options for defining fields and selecting edit checks
through point-and-click operations, and it contains options for defining range
checks, table lookup values, and check-digit validation criteria.
Developers can customize their own development environment and create
multiple applications and formats (data entry screens) simultaneously.
They can also create format specifications for such elements as image
backgrounds, field labels, and fields, and then save the specifications as
templates.
Data Entry Runtime Module

The Runtime Module ensures data integrity through validation during each phase
of data entry. Special features include immediate error detection and flagging,
re-key verification, data retrieval and validation from lookup tables.
DLL Support
Entrypoint provides full DLL support. Designers can create dynamic link
libraries for bringing C or C++ user procedures into their applications.
26
Host compatibility
Entrypoint applications can accept data in the best order for efficient data entry,
and then automatically re-format completed batches for compatibility with
mainframe databases and applications.
Applications
Designers can create a multitude of business applications. Here are just a few
examples:
Billing / Invoicing / Ordering
Human Resources / Employee Benefits
Accounting / Payroll
Student Registration
Document Tracking
Tax Return Preparation
Litigation Support
Inventory / Sales Analysis
Customer Information
Technical Specifications
The minimum hardware and software requirements for Entrypoint are: 80386
CPU, 8 MB memory, 40 MB hard disk (which includes space for typical user
applications), one floppy disk drive, an SVGA monitor, MS-DOS Version 3.0 or
later, and Microsoft Windows Version 3.1 or later (in 386 enhanced mode). An
SVGA monitor with 1024 x 768 resolution is recommended to use imaging
features.
Complete Product Support

Phoenix Software International backs Entrypoint Family of products with a full
range of support services, including a toll-free hotline, web site, training
programs, and design consultation.
For details about the MS-DOS version, go to the Entrypoint 90 page.
27
Entrypoint Plus
Entrypoint Plus, which is available for use in the Windows operating

environment, is a complete application development package for creating high-
performance data entry applications. Entrypoint Plus consists of an Application
Developer, a Data Entry Runtime Module (workstation), and an Administrator
Module.
Entrypoint Plus interfaces with standard communications products and
commercial networks.

Entrypoint Plus allows users to quickly and easily create a wide variety of data
entry applications with built-in validation, editing, and export options.
Entrypoint Plus contains many options for defining fields and selecting edit
check through point-and -click operations, and it contains options for defining
range checks, table lookup values, and check-digit validation criteria.
Developers can customize their own development environment and create
multiple applications and formats (data entry screens) simultaneously.
They can also create format specifications for such elements as image
backgrounds, field labels, and fields and them save the specifications and
templates.
Data Entry Runtime Module

The Runtime Module ensures data integrity through validation during each
phase of data entry. Special features include immediate error detection and
flagging, re-key verification, data retrieval and validation from lookup tables.
Applications
Designers can create applications for a multitude of business applications. Here
are just a few examples:
Clinical Trials
Litigation Support
Document Tracking
Customer Information
Accounting / Payroll
Employee Benefits
Inventory / Order Entry
28
Special Features for Clinical Trials

Audit Trail facility allows you to keep a record of all changes made to data
after the first entry, and comment as to why the change was made.
Application Revise Facility allows you to update a protocol during an
ongoing study and reformat batches after changes are made to an application.
Batch compare allows for independent double-entry verification by
comparing two batches created from the same application. It generates a
report of the differences between the batches.
CRF Template is a typical application for Clinical Trials with information on
demographics, medical history, examination, adverse reactions, and lab
information.
DOS to Windows Converter is an option to the many users of Entrypoint 90
to convert their applications to Windows.
Powerful Procedure Language

A powerful procedural language supplements Entrypoint Plus’ customization
potential, enabling developers to create custom edit routines, automate batch
external programs and procedures, instead of using DLL’s.
Administrator Module
Using this module, administrators can define operator profiles and security
features. They can also check the status of users logged on to a network.
Host Compatibility
Entrypoint Plus applications can accept data in the best order for efficient data
entry and then automatically reformat completed batches for compatibility with
mainframe databases and applications.
The minimum hardware and software requirements for Entrypoint Plus are:
Pentium processor
32 MB memory
1 GB hard disk space (which includes space for typical user applications)
One floppy disk
SVGA monitor
Windows 95 or later
1024 x 768 resolution on the SVGA monitor is recommended
29
Entrypoint 90
Entrypoint 90 is a complete software package for creating high-performance

data entry applications. The development system features a powerful full-screen
editor for designing data entry screens and comprehensive menus for selecting
field edits and error checks.

Menu driven Entrypoint 90 reduces development time and simplifies the design
process. Entrypoint 90 is designed so that non-programmers can create efficient
data entry systems.
Powerful Procedural Language

A powerful procedural language supplements Entrypoint 90’s customization
potential, enabling developers to create custom edit routines, automate batch
access functions, and call external programs and procedures.
Host Compatibility
Multiple export options and a communications bridge expedite the transfer of
data from the PC to the mainframe. Entrypoint 90 applications can accept data
in the best order for efficient data entry, then automatically reformat completed
batches for compatibility with mainframe databases and applications.
Key-To-Disk Replacement
A perfect replacement for outdated key-to-disk systems, Entrypoint 90
transforms the PC into a high-performance data entry workstation. Descriptive
field labels, timely prompts, customized error messages, and help screens walk
the operator through each task. Validation and editing functions ensure data
accuracy on a field-by-field basis.
Applications
Entrypoint 90 has the built-in functionality for a multitude of business
applications. Here are just a few examples:
Accounting / Payroll / Inventory
Billing / Invoicing / Customer Information
Human Resources / Employee Benefits
Document Tracking / Market Research
30
Tax Return Preparation

Oil/Gas Production
Clinical Trials
Student Registration
Litigation Support
Complete Product Support

Phoenix Software International backs Entrypoint 90 with a full range of support
services, including a toll-free hotline, web site, training programs, and design
consultation.
Entrypoint 90 runs on IBM PC/XTs, PC/ATs, PS/2s, and compatibles that run
MS-DOS Version 2.0 or later. The recommended hardware configuration for a
standalone PC is a hard disk with at least one floppy disk drive. The LAN
module workstation may run on a diskless PC. The development system requires
512K of free RAM, and a data entry workstation requires 320K. In standalone,
LAN, and distributed environments, IBM-compatible personal computers
provide a responsive, cost-effective platform for developers and end users.
For details about the Windows version, go to Entrypoint or Entrypoint Plus

(with special features for clinical trials).
31
Phoenix’s FALCON Family: FALCON32 for Windows

PC/FALCON for DOS
FALCON for Mainframe
FALCON32
Thirty-two bit processing has clearly improved the way PC applications handle
data. Before Falcon32, these improvements were not fully realized by the data
entry community. Other data entry products ran under Windows, but were still
limited by the old 16-bit processing platform.
Falcon32 has the power, dependability, and ease you expect from FALCON,
with the flexibility of Windows!
Whichever Windows operating system you are using, Falcon32 takes advantage
of the extra speed, reliability, and stability of 32-bit processing. Batches and
program functions are now processed faster than ever before, and you have the
ability to access any identified printers.
You can also take advantage of the power inherent in newer PCs. Not only do
you save on operation time, you also encounter fewer terminal lock-ups caused
by the PC’s inability to handle large data files.
Pop-Up Menus
Falcon32 provides easy access to program functions and commands through
pop-up menus. As with many other Windows applications, menu options vary
with the current display window or highlighted field. Menu selections are made
with a click of the mouse or by using shortcut keys.
Command Bars
Program commands are easily assigned to function keys or issued from the
application window. Falcon32 gives you a choice of using a PC/FALCON-style
command line or a movable command toolbar that includes a scrollable list of all
applicable commands. You can issue command functions without taking your
finger off your mouse!
TCP/IP Support
Falcon32 provides TCP/IP support to speed up data transfers and eliminate the
need for emulator software.
32
Pop-Up-Help
Falcon32’s comprehensive help includes three ways to find what you need. You
can choose between a table of contents or an index or you can search for a
particular word. Hypertext links quickly access related topics, and step-by-step
instructions help you perform most data entry tasks. Context-sensitive help
instantly displays details related to a specific item in a window, information, or
dialog.
FALCON Fits Your Environment

Whether you conduct data entry on the mainframe, PC, individual workstations
in a LAN configuration, or on all three, there is a flexible FALCON solution for
you. Basic FALCON operations are the same no matter what operating system
you’re using, so it’s easy to move between platforms. It’s also easy to grow and
change when you rely on a product that is both versatile and in sync with
industry developments. If you use Windows 3.1 and want to upgrade to
Windows 95 or later, you can continue to use Falcon32, without the hassle of
upgrading to a new version.
Falcon32 Uses Tried and True FALCON Methodology

The best thing about Falcon32 is its tried and true technology. Attention to the
needs of its users made FALCON their number one data entry product! With
Falcon32, you get the experience and innovation of over 16 years in the data
entry industry. No similar product can compare with the usability and security of
FALCON. You also get the best and most reliable data entry software available,
enhanced by the added capabilities of Windows.
33
PC/FALCON for DOS
PC/FALCON offers the same easy-to-use data entry features, data capture and
imaging capabilities, and high-level security as mainframe FALCON, and works
independently or with a host computer.
PC/FALCON is operable with or without Windows™ and includes a new pop-
up format-building menu that lets you define new formats in seconds.
Key features of PC/FALCON:

Formats and batches are created and maintained without using valuable
mainframe resources
Uploading and downloading are as easy as touching a key
PC technology makes possible key-sensitive, pop-up HELP screens, quick
data validation, and the ability to continue data entry operations even when
your mainframe is down
You can remap your keyboards to match your dedicated data entry keyboard,
making the transition to the PC painless
34
FALCON for Mainframe
FALCON has provided data entry technology to MVS and VSE mainframe
users since 1983 and has continued to grow and change since then. Today,
FALCON data entry software attracts many data entry users from different
software products that have not kept up with the times.
FALCON, the most powerful on-line data entry system, is also the easiest to
learn and operate. FALCON lets you define data entry screens and output record
layout in one easy step, and accommodates keyboard and keyless entry devices.
Applications are built by simply listing formats from anywhere within
FALCON.
Key Features of FALCON

All commands and functions may be assigned to PF keys
Easy edit commands perform data manipulation and validation without
coding
Includes extensive on-line help
24-hour technical support is available
Offers enhanced security and system resource conservation
Runs under z/OS, OS/390, VSE/ESA, and VM/ESA operating environments
Mainframe FALCON runs under CICS and as a standalone VTAM
application
For details about the Windows version, go to Falcon32; for details about the
MS-DOS version, go to PC/FALCON.
35
Viking Data Entry Systems
ImagEntry
ImagEntry® allows rapid and accurate keying from images of forms.

ImagEntry is typically used for indexing images and lifting data from image
documents such as medical claims and credit applications.
It is the only data entry system for keying data
from large format images like blueprints.
ImagEntry also complements ICR/OCR/OMR recognition engines. The results
file can be imported into ImagEntry, where field edits and double key verify can
be used to detect and repair errors. ImagEntry is today the most powerful
software system for keying data from images.
Application developers and system integrators can use ImagEntry to quickly add
data entry objects to image applications. An abundance of user exit subroutines
make it very easy to customize ImagEntry to meet the most demanding
applications. Designed to be integrated with other components, ImagEntry does
not contain scanner modules, ICR/OCR/OMR capability or other components
supplied by numerous vendors. Consequently, you can build superior
applications using the best components available.
ImagEntry supplements ICR/OCR/OMR data capture systems with extremely
productive and effective methods to key unreadable characters.
Key Features of ImagEntry:

“Paint” your form on screen – no programming experience needed
Adjust image views (zoom, up, down, left, right) during development
and production “on-the-fly”
Customize your forms with elements that increase entry accuracy,
reduce keystrokes and automate entry processes
Assign powerful field validation routines using menus and an entire
library of pre-defined routines
Works with most standard image formats, such as TIFF, PCX, CALS, etc.
Can run independently of the DBMS or file system containing images
and data
Handles multi-page image files
36
VDE
VDE is a Viking Data Entry System (VDE™) that since 1980 has been the
standard for multi-platform keyboard data entry from paper documents.
databases, spreadsheets or the indexing module that comes with most imaging
systems.
VDE offers traditional batch heads-down data entry from paper documents.
VDE has been widely used for conversions from key-to-disk machines such as
Nixdorf, Tartan, 4-Phase, 3741, DEF/2 and others. It runs on most common
operating systems, such as Windows, DOS, UNIX, VMS (AXP and VAX), and
others.
VDE Version 4.2 transforms your computer into an efficient workstation to key-
enter data rapidly and accurately, without additional special hardware. VDE
offers the 4GL tools you need for your most demanding key-entry applications.
It combines the best features needed for both “heads up” and “heads down” data
entry in an integrated system available on a variety of computer platforms: PC,
IBM RS6000, Sun, etc.
VDE System Capabilities:

Data Entry Application Generator
VDE’s 4GL tools make it easy to develop applications that reduce time and cost
expenditures.
Forms Development Facility
The Viking Forms Painter is used interactively to create and modify screen
“Forms” that define data entry jobs. The full-screen system is so intuitive and
easy to use that end users and data entry operators can design and develop their
own jobs.
File Reformatting
General purpose 4GL reformat capabilities let you take advantage of new
technology and techniques to increase data entry productivity, while at the same
time produce output files that meet the specifications of existing systems.
ISAM File Support
VDE provides fast on-line access to keyed Indexed Sequential Access Method
(ISAM) data files and tables.
37
Customization
VDE is an open system. Using the optional API and associated tools, VDE is
easily customized to meet requirements of the most demanding applications. Use
standard programming languages like COBOL, FORTRAN and C – no need to
learn specialized “data entry languages.”
Menu System
Viking makes wide use of drop-down menus which are easily customized for
unique needs. VDE can even launch other programs, provides additional security
and maintains an audit trail.
Key Features of VDE:

Full-screen, “fill-in-the-blank” screen forms
Intelligent double-key verify mode with immediate correction capability
Insert, delete, modify and append data records
99 Program levels or record types
Powerful, easy to use file search
Create, update, re-key verify into ISAM files
File conversion, reformatting utilities
Table lookups
Automatic balancing options
VDE Applications:
Health Claim Forms, Clinical Trials
Wholesale Lock Box, Market Research
Engineering Design, Inventory Control
Populate Databases, Off-load Mainframe
Survey Forms, Tax Forms
Accounting, Payrolls
38
VDE+Images
VDE+Images combines the best features of Viking’s ImagEntry and VDE data
entry software, allowing you to key from both paper and image documents with
one software package.
VDE+Images System Capabilities:

Open Architecture Integrates with Existing Systems
VDE+Images has an open architecture and flexible design that makes it easy to
integrate with any document imaging system. VDE+Images can be readily
customized using the optional API. Custom field edits and exit routines are
coded in the industry-standard C language.
Pre-existing Data
VDE+Images can import pre-existing data to form the basis for data records.
This data can come from the scanner, bar-code readers, MICR readers,
recognition engines (OCR, OMR and ICR) and other data sources. For example,
the MICR data from checks can be imported into VDE+Images for wholesale
lockbox applications.
Power/Precision Data Entry
VDE+Images has all the features of the Viking Data Entry System that are
required for high speed, high volume, high accuracy keying. Viking’s customers
traditionally key data at rates well above the average power/precision keying
speed of 11,600 keystrokes per hour. Viking’s customers demand data accuracy
far in excess of the capabilities of today’s recognition engines.
Data Capturing from Unstructured Paper or Image Documents
VDE+Images is specifically designed for those imaging applications where the
location of the data is not predictable, either because the documents do not have
a standard format or because there are a variety of forms that arrive in a random
order, such as wholesale lockbox and land title records.
Multiple Relationships between Images and Data Records
Most imaging systems automatically generate one record for each image.
VDE+Images is unique among key-from-image (KFI) products in that it allows
data records to be stored in the most useful and logical way. So you can have
data from any number of images on one record, put data from one image on
multiple records, or spread data from several images across several records –
whatever works best for your application.
39
Image Manipulation
VDE+Images is designed to allow the operator to quickly and easily manipulate
the image in order to locate and key the appropriate data. Hot keys let operators
zoom, move, rotate and invert images with a single keystroke. They can also
quickly switch from one “zone” of the image to another.
Image Sets
Many imaging applications need to group images into sets that represent the
pages for a multi-page document, or know the beginning and ending images of
file folders. VDE+Images has features to define the beginning of an image
set and to advance to the next set.
Hardware Requirements:
A basic PC with Windows 95, 98, 2000, ME or Windows NT is sufficient for the
client workstation.
High powered workstations are unnecessary. VDE+Images will work with any
network, including Windows NT and Novell. The size of the monitor needed
depends on the applications. Many jobs work fine with a basic 15” monitor. A
17” or 19” monitor may be required if the entire document needs to be displayed
at full size or larger.
40
Frequently Asked Questions
What are the hardware requirements for the latest version of VDE?
Viking Data Entry (VDE) version 4.0 for Windows 95, or Windows NT, requires
10 MB of hard disk space for VDE, plus space for data files, sequential and
indexed tables on the server, and none on the client. VDE requires only
minimum memory requirements on the Workstation and no additional memory
requirements on the server.
I receive a key-from-image module free with my document imaging system.

Why should I spend my money for ImagEntry?
If it is free, your vendor probably knows what it is worth. ImagEntry pays for
itself in three ways:
(1) You will reduce your key entry costs through greater productivity
(2) You can set up applications quicker
(3) Your data will be more accurate
Do I need data entry if I already have the images in my computer?

Remember, an image is just a picture, a collection of pixels that represent the
physical document. Before a computer can do anything with the information on
the document it must be converted to ASCII data (i.e. letters and numbers, not
pictures of letters and numbers). There are two ways to convert image
information to ASCII:
(1) Manually key enter the data using a key-from-image module like ImagEntry
(2) Use a recognition engine (OCR, ICR or OMR)
What are the system requirements for ImagEntry?

You don’t need a high-end system to run this high-end software. ImagEntry is
designed to run with minimal requirements. The minimum System Requirements
is a 486 processor, with 8 MB RAM, and Windows 95 or Windows NT. If the
documents are unstructured and free form, you probably need a 17” monitor.
Monochrome is fine for black and white documents and most documents that
have some color on them.
Why do I need ImagEntry?

I am getting a recognition engine that will eliminate all manual data entry.
If you believe that any recognition engine is 100% accurate I want to make you a
deal on some beach front property in Arizona!
41
You will always have some unreadable characters to be hand entered. Did the
salesman tell you about substitution errors? How are you going to find them and
fix them? ImagEntry can do it.
Tell me about data accuracy - my OCR system is pretty good.

How do I compare it?
Manually keyed and verified data averages one character error in 10,000 or
99.99% accurate. OCR salesmen promise to read 95% to 99%. This means they
can't read from 100 to 500 characters out of 10,000. Worse yet, the substitution
error rate can be really bad.
What type of technical support is available for ImagEntry?

An online manual, a Getting Started manual with tutorial, and complete Help
System is packaged with ImagEntry. Additionally, we offer Viking’s unique
Field Help facility for all technical support issues – contact by phone, fax and
email. This includes:
New Releases
Product Updates
Technical Questions
Extended support, our Annual Maintenance Contract , may be purchased for
18% of the license price.
Viking’s customers enjoy unsurpassed level of quality in the industry. Our goal
is for a Customer Care representative to answer each call as it comes in and
respond immediately with a resolution to your issue.
What do you mean by substitution errors?

Recognition engines make two kinds of errors, Rejects and Substitutions. A
Reject is a character that it couldn’t read, and it recognized that it couldn’t read
it. Substitutions are characters that are misread. For example, it interprets an “e”
as a “c”.
Why does ImagEntry clutter up the screen with so many data fields?
Your competitor simplifies it by only showing one field at a time.
Ask our competitor to talk to you about keying speed and rhythm. If they think
keying rhythm is some new kind of music, run, don’t walk, to the nearest exit.
42
We need to key data from images of large drawings in the CALS format.
Can I do that with ImagEntry?
Certainly! Define image snippets that display the information to be keyed. You
can move and zoom the image to find information that is not in the expected
location.
How do I know how fast our people key data?

If your application doesn’t measure and report some metric like key strokes per
hour or records per hour, it probably means you are not keying very fast. We
hear horror stories about keying rates that are only 10% of the national average
of about 11,600 key strokes per hour. You need to understand why you key at
your present rates. Do you?
Why can people key faster with Viking products?

To key rapidly the applications must do three things: (1) Eliminate unnecessary
keystrokes, (2) Promote rhythmic keying, and (3) Minimize hand movements.
Viking has been creating the tools and techniques to eliminate keystrokes since
1980. Many of these techniques are known only to those people who specialize
in power/precision data entry. Superstar key entry operators regularly average
more than 15,000 keystrokes per hour by building a smooth rhythm in their
keying. They cannot achieve these speeds with applications that have short
bursts of keying that leads to herky-jerky rhythm. Clearly, if you have to take
your hand off the keyboard to use a mouse you lose many keystrokes.
What kind of keystroke average should I expect?

Keystroke speeds are very application dependent and must be considered in that
context. However, in 1989 the Data Entry Management Association reported that
the USA national average for all applications was about 11,600 key strokes per
hour. This is a mixture of a lot of different types of keying jobs. If you are only
keying numbers you should expect to do much better. If you key alphabetic
letters it may not be so easy to reach the average speeds. If your applications
requires you to Athink@ before keying, or if you have to search for the data it
will be much slower. And of course there is a big difference between
individuals. The important thing to look for is whether the keyer ever has to wait
for the system. Many applications can be improved dramatically by improving
the design to increase keying speed and eliminate keystrokes.
43
Why the big emphasis on key strokes per hour?

It is a common metric but is not always the best one to use. After all, the real
measure of productivity is how long it takes to process a given number of
transactions or data records.
I cannot afford to buy new hardware and software.

It just is not in our budget. What can I do?
You need to look at the total costs. The rental and maintenance for the hardware,
the mainframe charges and the total labor costs for your data entry applications.
If you are using dedicated key to disk machine, or keying directly on-line to a
mainframe, the chances are pretty good that you can reduce you overall data
entry costs with a modern system. When you identify all the costs involved you
may be surprised how much money you have available that only needs to be
reallocated.
Under what platforms does VDE operate?

VDE operates with the following operating systems: Windows 95, Windows NT,
MS/DOS, VMS, UNIX (most popular systems, including IBM, DEC, HP, Sun,
Intel, SCO, etc.).
Do you provide training on VDE?

Yes! Training classes are scheduled in Tulsa, USA. VDE training is a 3-day
session. You also receive a tutorial in the complete set of manuals.
Contact us at info@mb-imaging.com for a current class schedule.
How can VDE help me process my information?

What can I use VDE for?
Our customers find a variety of uses for VDE. It makes keying data faster, more
accurate and requires little overhead. While each business purpose is different
and unique to that particular business, here are some of the processes for which
our customers use VDE:
Keying data from paper forms
Credit Applications; Invoices
Insurance claim forms
Air Bills and Way Bills
Land records; Survey Forms
Litigation Support documents
44
Evaluating Data Entry Systems
Imaging service bureaus often have opportunities to do data entry from the
images they scan for their clients. This is an important source of additional
revenue. Most firms start by simply indexing images. Larger revenues come
from entering content to be input into the client’s data processing systems.
Data entry is a double-edged sword for the service bureau. First, it is an
opportunity to increase billings and to get scanning work from clients who prefer
a one-stop service. On the other hand, when the service is not managed properly,
it can become a big cost and a money-losing proposition. At best, most imaging
bureaus fail to achieve the profit potential of their data entry services.
The key to profitable data entry is the data entry software system you select. You
must have a system that maximizes the productivity of your data entry operators.
Anything less adversely affects your bottom line.
Data entry labor costs overwhelm the software cost. For example, assume your
fully burdened labor costs are $25,000 to $30,000 per year. When you achieve
only a 5% gain in productivity you will save up to $1,500 per year, which is
more than the cost of professional quality data entry software. Most companies
see much higher productivity gains than a mere 5%. And these savings go on
year after year.
There is also a labor cost to set up and maintain data entry jobs. If this is a time
consuming task requiring programmers and other expensive labor, your
profitability will be further eroded. If you can not react quickly to opportunities
you will not get the business. So, it is imperative that your new data entry system
provide for quick and easy job setup and maintenance. Clients know what data
entry costs. You cannot overcharge in order to make up for your inefficiencies.
Competition will kill you.
Operator Productivity
I am amazed at people who buy the best scanners because they understand the
value of efficiency and reliability, but don’t buy good data entry software. Avoid
using amateur equipment to do a professional job. I often hear, “Why spend
money for data entry software when I already have MS Access?” or “I can whip
out something with Visual Basic.”
45
The problem is that the average productivity rate for professional data entry
operators is 11,600 keystrokes per hour. It has been reported that the average
operator entering data with Access or standard Windows panels is only about
1,500 keystrokes per hour. This 9 to 1 productivity improvement quickly pays
for good data entry software.
Ask yourself these questions:
How fast are your operators?
Do you even measure keying speed?
What is your error rate?
What does it cost to find and correct errors?
An old saying that is particularly true of data entry is, “If you can’t measure it,
you can’t fix it.”
Data Accuracy
Inaccurate data causes the client to take their business elsewhere. A single data
error may cost the client hundreds, and even thousands, of dollars to repair. The
intangible costs in good will and customer service due to data errors is huge.
This requires a different mind set from the quality control for scanning. If you
mess up an image you can just re-scan it for the client, no big deal. If you make a
data error, it may very well be a big deal.
Professional quality data entry software systems offer a full range of data
validation and error detection features.
Professional Quality Software

Professional quality data entry software, which has been available for decades,
addresses these essential issues of productivity and accuracy. However, the top
vendors have continuously enhanced their products to support the latest advances
in imaging. As a service bureau, you need to search for software that meets your
exacting requirements.
Features and Functions

What are some of the features you should look for? And how do those features
affect profitability?
Technique vs. Technology
Profitable keyboard data entry involves using many techniques that have been
developed and proven over a long period of time. High tech solutions have less
46
of an impact here. The best systems apply and use many small factors to build
substantial increases in productivity.
Ergonomics
Production keyboard data entry uses function keys for specialized features and
functions. The whole idea is to minimize hand and finger motion to allow
operators to achieve their maximum potential keying speed. It has been proven
that the fastest keyers have the best techniques and make the fewest errors.
Image manipulation
Newer data entry systems have features and functionality that let your operators
easily manipulate images so that the data can be key entered. For example,
efficient moving, zooming, etc., is done with hot keys and not a mouse. The next
image should appear automatically and instantly after the last data item is keyed.
Enter vs. Tab
The big Enter key is under the right hand and is the traditional key typists and
data entry operators use to complete fields or lines. The Tab key is needed to
advance over data entry fields to some predetermined field. It is a mystery why
Microsoft thinks the Tab key should be used to complete fields.
Field navigation
Look for hot keys to go to the next and previous fields, and to the next and
previous images. Moving the hands from the home keys to use a mouse is a
significant slow down.
Eliminating keystrokes
Features that automatically populate fields with data will reduce keystrokes and
increase productivity and accuracy. A special key to duplicate previously keyed
data is required.
Skipping optional fields
It should be simple, even automatic, to skip over fields that normally are usually
empty.
Instant data validation
Data should be validated as soon as it is entered. Only the right kind of
characters should be accepted in a field (e.g., no letters in numeric fields). This
validation must be accomplished at the time the key is depressed, so errors can
be corrected while the eye is on the data and the finger is on the key.
47
Extensive field edits

Date validations, range checks, table lookups and calculations are just some of
the many edits that can be performed on each field as it is completed. It is crucial
to detect errors at the earliest possible time.
Error messages
Errors should lock the keyboard and require the keyer to acknowledge the error
message and correct the error.
Double-key verify
The time-proven data entry method to ensure data accuracy is to key the data
twice, preferably by different people. This process results in 99.99% accuracy.
Good data entry systems make this process simple and fast, yet it must be
optional for those projects where double keying is not desired. Also, it is
important that error correction be easy and safe. Verify error detection must be
done at the keystroke level.
Statistics reports
In order to successfully manage the operation you need good reports on operator
productivity and errors. The more information you have, the better able you are
to allocate costs and ensure that jobs are quoted accurately and performed
profitably.
In summary, data entry can be far more profitable for most service bureaus than
it is now. When you use the right tools and time-proven data entry techniques
you see significant increases in productivity and accuracy. Professional Data
Entry Systems are the key to success for service bureaus.
48
Conversions:
What are the Issues for Data Entry Systems?
Data entry systems tend to have a long life. When the time comes to change to a
different system it is usually a new experience for the staff involved in the
project. The main objective in this article is to suggest some ways to plan the
conversion so it will be as smooth and painless as possible.
Data entry systems are acquired for many reasons. Today we see many old
systems being changed because they are not Y2K compliant. The old systems
may be on a mainframe that is being downsized or the applications may be
moving to a network of PC’s.
Old dedicated data entry systems such as Nixdorf, REI Tartan, Entrex, CMC,
Motorola Four-Phase, Pertec, Honeywell, CADE, Mohawk, IBM 3741, and
others are being replaced because they are becoming unreliable, difficult to
maintain and extremely expensive to repair and operate.
Goals and Objectives

The first step is to determine the goals and objectives of the project. Some of the
questions to ask are:
When can we start?
When does the conversion need to be complete?
What needs to be improved and changed?
Do we want the new system to closely resemble the old one, or do we want
to re-engineer the jobs to maximize productivity?
Information Gathering
Once goals and objectives have been established and approved, you need to
gather the appropriate information.
You need to identify each data entry job and determine if it is to be converted or
if it is obsolete or redundant. For each job you should count the number of
screens and classify them as being simple, complex or duplicate.
You will also need to gather the keying instructions, output formats, sample
source documents and editing criteria for each field.
49
Estimate Time Requirements

Once all the information has been gathered and tabulated, you can estimate the
time to convert screens to the new system. Develop estimates for the three
categories of screens: simple, complex and duplicates.
You can then estimate the time to convert the screen formats by multiplying the
various estimates by the number of screens in each category. Automated and
semi-automated conversion tools can sometimes be used to expedite this process.
Next, estimate the time to add the proper data edits and validations. This will
depend on the editing and validation facilities of the new data entry system.
Finally, make estimates of the time to create the necessary documentation for the
new system. It is important to do the documentation at the same time the
conversion is taking place or else the odds are very good that it will never be
completed. Good documentation will pay for itself in reduced training time,
reduced errors and reduced administrative and supervisory time. Remember,
time is money!
Once the estimates are complete you can determine how many people will be
required to complete the conversion within the time frame specified in your
goals and objectives.
At this time you can factor in the cost to re-engineer some or all of the jobs to
achieve improved productivity. A small percentage decrease in data entry costs
will continue to accrue year after year and is usually well worth the initial cost.
Team Members
The conversion team should consist of the appropriate people from the data entry
group, the systems and programming group and your vendors and professional
services suppliers. Different skills are needed for different aspects of the project.
Availability of the appropriate people will be key to meeting the schedule.
Training
Training may be required for the development staff and the data entry operators.
Training may be provided by simply reading manuals, by using internal trainers,
or from your data entry software vendor. In most cases a combination of these
sources will be the optimal solution.
The amount of training often depends on the schedule. If it is tight you may want
early training for your developers to jump-start the conversion process.
50
Operator training is usually done in-house by the developers and supervisors. It

should not take more than a couple of hours.
In-House vs. Out-Source

The decision to use internal staff for the conversion or to out-source the work
depends on many factors. Your organization’s policy and culture in this regard is
the principal factor.
Availability of staff is another issue. Often internal staff is limited or not
available in the desired time frame.
Outsourcing to the software vendor will usually result in a quicker conversion
because of their expertise and experience. When you include fully burdened
costs for internal staff the actual cost will probably be about the same.
An advantage of the in-house approach is that you then have the expertise to
maintain your system in the future. Of course, that assumes that the staff that
does the work will be available in the future.
If you have made a wise selection of the data entry software system you should
expect that overall productivity will be at least as good as your present system. If
you are able to take advantage of faster systems with newer features and
technology then your operators should become more productive.
People who key data all day long are very sensitive to the keyboard. A slight
change in the touch and feel will be noticed by your fastest operators. There is a
definite adjustment period that you should plan for.
Productivity will follow the traditional S-shaped learning curve. When you first
switch over to the new system the operators will be slower at first. As they
become familiar with the new system they will gradually return to the same, or
even a higher, speed. This learning curve effect can take a few days to a couple
of months.
You can expect initial resistance to the new system if the old system was
satisfactory. This is human nature: we are all resistant to change. Some people
adapt more quickly than others do. Psychological factors have a lot to do with
the learning curve.
If operators approach the new system with a positive attitude, things will
progress faster and smoother.
51
It is important to prepare them for the change and show them how the new
system is going to be better. Make it a fun challenge rather than something
“they” have mandated from on high. Involve them in the process and solicit their
opinions on how to improve jobs. Operators who feel an “ownership” in the new
system adapt readily and happily.
Ongoing Maintenance
Someone on your staff should learn to set up new data entry jobs and make
modifications to existing jobs in the future. It is not good to be at the mercy of a
vendor or outside consultant for every little change that you need.
It has often been said that “change is the only constant in life”. This is certainly
true of data entry jobs.
Preparing for Key-from-Image

Document imaging is becoming a mainstream technology. Even if you have no
plans for imaging today, the odds are very good that you will be looking at
imaging projects in the foreseeable future.
Data entry from document images is slightly faster than keying from paper
documents. Plan ahead, the best data entry software systems allow you to easily
migrate to key-from-images. There are things you can do when you convert your
data entry jobs that will prepare you for the future. Investigate this with your
software vendor.
Viking Software’s Conversion Experience

The staff at Viking Software Solutions has assisted in the conversion of the
following data entry systems:
Key Master CADE Falcon S1DES

Nixdorf Series 1 Entrex REI Tartan
DPS Sovereign Mohawk Four-Phase
MDS Life Works Data 100
52
Keying from Images when your

Documents are not Forms
Some imaging applications require that large amounts of data be entered from
documents that are not standardized forms. Forms processing is a well
established technology for imaging systems, but it assumes the documents are
highly structured. What about those applications where the documents are
unstructured, or freeform. Traditional forms processing technology doesn’t work
well, if at all. The many benefits of imaging can still be achieved, but special
attention needs to be paid to data entry in these cases to keep from causing
system bottlenecks.
What characteristics do these documents have?
Unlike structured documents, or forms, the data to be keyed is not in the
same place on each page of the unstructured document. Often it is not even
on the same page.
Structured documents are usually the same, or when there are multiple pages
they appear in a consistent and predictable manner.
There may be many different kinds of data to be keyed from any page. There
may not be any data, or there may be lots of data, on any given page. In
general, the document determines what is to be keyed.
Unstructured documents are not amenable to the successful use of
recognition engines (OCR/ICR/OMR) because the zones of interest are not
definable. Recognition is seldom successful on unstructured documents
because the zoning problems and laborious job setup. Recognition rates are
poor when the zones are not precise.
What type of documents fall into this category?

Land Records
General purpose image indexing
Litigation Support
Wholesale Lockbox
Many Accounts Receivable Applications
Accounts Payables
Credit Card Vouchers
53
When implementing an imaging system, what data entry issues need to be

considered for this type of image?
Then for each issue, what characteristics should you look for in key-from-image
data entry software/hardware if you deal with this type of images?
Often there are a variable number of pages (images) that make up a
document set (think of the document set as a file folder containing a variable
number of pages). The data to be keyed is associated with the document set,
not necessarily the individual images. Thus, there can be a many-to-many
relationship between data records and images.
Often there is a need to view the images in a document set and to advance to
the next document set. This requires special functions to advance to the first
page of the next document.
Balancing financial documents is often required.
Unstructured documents usually require 19” or larger monitors for
productive keying because the full image must be displayed. Structured
documents can be keyed using small monitors because only a snippet of the
image needs to be displayed in order to key one or more data fields.
Interfaces with other parts of the imaging system.
Compatibility with any key-from-paper systems that will continue to work in
parallel with the imaging system.
Conversion and training.
Ease of maintenance – making changes to the job setup.
In conclusion, if your organization is implementing imaging and you deal with

this type of document, don’t let data entry problems keep your imaging project
from being successful. A little attention to the issues outlined above will ensure
that you put together a complete system that fits the way your organization
works.
54
Top 10 ways to lose Money doing Data Entry
Cost Effective Data Entry from Images

Organizations with document image systems must do data entry from the
scanned images of the documents. In the simplest case, the images need to be
indexed for retrieval. In most applications there is a need to lift data from the
images to be input to other programs and systems. The cost of this data entry can
be excessive if the wrong data entry systems are employed. This often leads to
unexpected labor costs that can eat up all the savings of the imaging system.
Let’s look at the factors involved and see how to avoid excessive costs.
The key to profitable data entry is the data entry software system you select. You
must have a system that maximizes the productivity of your data entry operators.
Anything less adversely affects your bottom line.
Data entry labor costs overwhelm the software cost. For example, assume your
fully burdened labor costs are $25,000 to $30,000 per year. When you achieve
only a 5% gain in productivity you will save up to $1,500 per year, which is
more than the cost of professional quality data entry software. Most companies
see much higher productivity gains than a mere 5%. And these savings go on
year, after year, after year!
Yet, every day we hear horror stories about how much more costly it is to key
input data from document images than when they keyed from the original paper
documents. Why does this happen?
There is also a labor cost to set up and maintain data entry jobs. If this is a time
consuming task requiring programmers and other expensive labor, your cost will
be excessive. Inability to be able to react quickly to new opportunities can cost
your company through loss of business. So it is imperative that your new data
entry system provide for quick and easy job setup and maintenance.
I am amazed at people who buy the best scanners because they understand the
value of efficiency and reliability, but don’t buy good data entry software. Avoid
using amateur equipment to do a professional job. I often hear, “Why spend
money for data entry software when I already have MS Access?”, or “I can whip
out something with Visual Basic.”
55
The problem is that the average productivity rate for professional data entry
operators is 11,600 keystrokes per hour. It has been reported that the average
operator entering data with Access or standard Windows panels is only about
1,500 keystrokes per hour. This 9 to 1 productivity improvement quickly pays
for good data entry software.
Ask yourself these questions. How fast are your operators? Do you even measure
keying speed? What is your error rate? What does it cost to find and correct
errors? The old saying, “If you can’t measure it, you can’t fix it” is particularly
true of data entry.
Data Accuracy
Inaccurate data is expensive to correct and often causes your customers to take
their business elsewhere. A single data error may cost hundreds, and even
thousands, of dollars to repair. The intangible costs in good will and customer
service due to data errors is huge. This requires a different mind set from the
quality control for scanning. If you mess up an image you can just re-scan, no
big deal. If you make a data error it may very well be a big deal.
Professional quality data entry software systems offer a full range of data
validation and error detection features.
Professional Quality Software

Professional quality data entry software, which has been available for decades,
addresses these essential issues of productivity and accuracy. However, the top
vendors have continuously enhanced their products to support the latest advances
in imaging. You need to search for software that meets your exacting
requirements.
Features and Functions

What are some of the features you should look for? And how do those features
affect profitability?
1. Technique vs. Technology. Profitable keyboard data entry involves using many
techniques that have been developed and proven over a long period of time. High
tech solutions have less of an impact here. The best systems apply and use many
small factors to build substantial increases in productivity.
2. Ergonomics. Production keyboard data entry uses function keys for
specialized features and functions.
56
The whole idea is to minimize hand and finger motion to allow operators to
achieve their maximum potential keying speed. It has been proven that the
fastest keyers have the best techniques and make the fewest errors.
3. Image manipulation. Newer data entry systems have features and functionality
that let your operators easily manipulate images so that the data can be key
entered. Efficient moving, zooming, etc. is done with hot keys and not a mouse.
The next image should appear automatically and instantly after the last data item
is keyed.
4. Enter vs. Tab. The big Enter key is under the right hand and is the traditional
key typists and data entry operators use to complete fields or lines. The Tab key
is needed to advance over data entry fields to some predetermined field. It is a
mystery why Microsoft thinks the Tab key should be used to complete fields.
5. Field navigation. Look for hot keys to go to the next and previous fields, and
to the next and previous images. Moving the hands from the home keys to use a
mouse is a significant slow down.
6. Eliminating keystrokes. Features that automatically populate fields with data
reduce keystrokes and increase productivity and accuracy. A special key to
duplicate previously keyed data is required.
7. Skipping optional fields. It should be simple, even automatic, to skip over
fields that normally are usually empty.
8. Instant data validation. Data should be validated as soon as it is entered. Only
the right kind of characters should be accepted in a field (e.g., no letters in
numeric fields). This validation must be accomplished at the time the key is
depressed so errors can be corrected while the eye is on the data and the finger is
on the key.
9. Extensive field edits. Date validations, range checks, table look-ups and
calculations are just some of the many edits that can be performed on each field
as it is completed. It is crucial to detect errors at the earliest possible time.
10. Error messages. Errors should lock the keyboard and require the keyer to
acknowledge the error message and correct the error.
11. Double key verify. The time-proven data entry method to ensure data
accuracy is to key the data twice, preferably by different people. This process
results in 99.99% accuracy. Good data entry systems make this process simple
and fast.
57
Yet it must be optional for those projects where double keying is not desired.
Also, it is important that error correction be easy and safe. Verify error detection
must be done at the keystroke level.
12. Statistics Reports. In order to successfully manage the operation you need
good reports on operator productivity and errors. The more information you
have, the better able you allocate costs and ensure that jobs are quoted accurately
and performed profitably.
In summary, data entry can be far less costly for most companies than it is now.
When you use the right tools and time-proven data entry techniques you see
significant increase in productivity and accuracy. Professional data entry systems
are the key to success for service bureaus.
58
How Will Our Data Entry Systems Help Solve Your Document Management Problems?
How Important is Accuracy?

Quick Launch
Getting more work done with fewer resources is key to survival these days.
Data Entry
Systems The amount of hand-whiting paper documents that are a day-to-day part
of your job is on the increase: invoices, checks, bills, orders, applications.
Free The list is endless. And each document must be reviewed, approved,
Downloads sorted, processed, and ultimately, filed and stored.
Take a look at the methods your organization uses for data entry:
The time-
time- How important is the impact of errors:
proven reprocessing costs, unhappy customers, lost business, etc?
methods of How much it costs your organization to detect and correct errors?
key-
key-entering
data result You know today’s ICR/OCR systems have unsatisfactory error
rates for most real-world documents.
in 99.99%
You know, there must be a better way.
Accuracy!
There is!
Our Power/Precision Data Entry Systems
MB-Imaging provides Viking Data Entry Systems – the most powerful today software systems
allowing rapid and accurate keying data from both paper and images:
VDE is a Viking Data Entry System (VDE™) that since 1980 has been the standard for
multi-platform keyboard data entry from paper documents.
databases, spreadsheets or indexing modules that come with most imaging systems.
ImagEntry is the power software for keying data from images of forms (structured
documents), designed to be the key entry component of high performance document
imaging systems. ImagEntry supplements ICR/OCR/OMR data capture systems with
extremely productive and effective methods to key unreadable characters.
VDE+Images combines the best features of Viking’s VDE and ImagEntry data entry
software, allowing you to key from unstructured documents (which are not amenable to the
successful use of OCR/ICR/OMR recognition engines because the zones of interest are not
definable).
Everything on Your Fingertips
Our Power/Precision Data Entry Systems, perfect for integration into any document imaging
applications, keep vast amounts of data at your fingertips, allowing you to index images and
key-enter data: rapidly, accurately and affordably!
Even if this offer is not right for you, since you have read thus far, something caught your
interest.
Why don't you call us today (or visit www.mb-imaging.com) for information about How Will
Our Data Entry Systems Help Solve Your Document Management Problems?
59
Related Links
http://www.all-for-one.de
http://www.captivacorp.com
http://www.cardiff.com
http://www.easy.de
http://www.imagepub.net
http://www.mb-imaging.com
http://www.microsystemsonline.com
http://www.phoenixsoftware.com
http://www.vikingsoft.com

60

Data Entry Secrets

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Data Entry Secrets

Uploaded by

Copyright:

Available Formats

Free Data

MB-Imaging, Delémont Switzerland

Our Power/Precision Data Entry Systems

Everything on Your Fingertips

Why you need Professional Data Entry Software 1

Why You Need Professional

How much data are you entering?

How much do you pay your employees?

How important is accuracy?

Character Level Validations

Field Level Validations

Record Level Validations

Double Key Verify

The Importance of Power/Precision

One of the often overlooked and underestimated elements of a Document

ICR/OCR Accuracy Issues

Reject Correction Costs

Key Entry Speed

Two Cost Factors

Substitution Error Costs

Manual Detection Methods

Correcting Substitution Errors

Miscellaneous: Other Factors that Influence Recognition Rates

Recognition accuracy depends on many factors. Good quality documents with

to do manual entry in the beginning until satisfactory document recognition

advances in technology and improved methodology. Are you willing to make a

Keypunch Style Keyboard

Enter Key vs. Tab Key

Modify Existing Data

Image Display Speed

Application Development Costs

Multiple Image Formats

Flexibility and Adaptability

Captiva Software Corporation

FormWare is designed to serve complex capture needs in higher-volume, multi-

Desktop is fully expandable, entry-level version of FormWare. It combines

InvoicePack uses innovative optical character recognition (OCR) and neural

Next, InvoicePack automatically attempts to identify the invoice type by

OCR for Forms

Data verification is a high-speed, heads-up procedure confirming all data entered

Applications (structured forms):

AnyForm Technology incorporated special features to process the unique

Applications (unstructured forms):

Phoenix Software International

Entrypoint Family: Entrypoint for Windows

Entrypoint is a complete application development package for creating high-

Quick, Easy Development

Data Entry Runtime Module

Complete Product Support

For details about the MS-DOS version, go to the Entrypoint 90 page.

Entrypoint Plus, which is available for use in the Windows operating

Quick, Easy Development

Data Entry Runtime Module

Special Features for Clinical Trials

Powerful Procedure Language

Entrypoint 90 is a complete software package for creating high-performance

Quick, Easy Development

Powerful Procedural Language

Tax Return Preparation

Complete Product Support

For details about the Windows version, go to Entrypoint or Entrypoint Plus

Phoenix’s FALCON Family: FALCON32 for Windows

FALCON Fits Your Environment

Falcon32 Uses Tried and True FALCON Methodology