Knowledge management refers to an organization's strategic efforts to gain a competitive advantage by capturing and using the intellectual assets

held by its employees and customers. Efforts to archive best practices and lessons learned, and to make better use of information stored in databases, also fall under the rubric of knowledge management. Advocates of knowledge management believe that capturing, storing, and the distributing knowledge will help employees work smarter, reduce duplication, and ultimately produce more innovative products and services that meet the customers' needs and offer a good value. If a company knows something (e.g., changing tastes of the customers, innovative solutions to international tax issues, or how to use information systems to better monitor production processes) that its competitors do not, then that company has an opportunity to offer a distinguishing product or service. Knowledge management, as a business practice, impacts the entire organization by helping employees, managers, and executives share information and best practices that positively impact collective performance. Unlike downsizing, which emphasizes the reduction and control of costs (often through attrition and layoffs), knowledge management is a value-adding practice that seeks to enhance profits, innovation, and decision making by providing more and better information to every member of the organization. To better understand why knowledge has become a critical factor in businesses, we need to understand that the United States and many other industrial countries are moving toward a knowledge economy. A knowledge economy is one where a majority of workers spend their day applying know-how to the production of goods and delivery of services. In a knowledge economy, employees work to improve decision-making, design, and delivery processes, while only a limited number of people are involved with the actual manufacturing of goods. Important questions that we might ask about a knowledge economy include:

How many people now spend their day applying knowledge? How did the change to a knowledge economy come about? How do organizations go about managing knowledge?

American labor trends indicate that the percentage of people working in an information-intensive capacity is increasing while the number of people working in agriculture, manufacturing, nonprofessional service industries is decreasing. As another indicator of the shift, during the second half of the twentieth century, knowledge-intensive companies (those that have 40 percent or more knowledge workers) account for 28 percent of the total U.S. employment and produced 43 percent of all new employment growth. The rapid increase in knowledge-intensive work is often attributed to communication technologies, and especially digital technologies, that allow employees to transfer or access large amounts of data in minutes. Since the end of World War II, the world has seen the invention of the first programmable computer, satellite technology, fax machines, microprocessors, floppy disks, portable computers, cellular telephones and pagers, and the World Wide Web. All of these technologies are historically important because they allow great quantities of information to be shared with partners who are geographically separated from us and who, using earlier technologies, might have had to wait hours, days, or even weeks to receive information. Technology has, in effect, brought people closer together by allowing voice, text, and images to be rapidly transmitted across great distances. In 1969, the Department of Defense launched the Advanced Research Project Agency, which created a distributed network (precursor to the Internet) that allowed researchers to share information and connect with other computers on the network. Later, researchers added e-mail bulletin boards to the system so that messages could be transmitted back and forth. This broad digital network took information sharing to an entirely new level. Whereas a fax machine might be able to transfer 2,000 words from New York to Los Angeles in a matter of minutes, this new digital networktoday represented by the World Wide Web allows information to be transmitted at the speed of light. Current statistics indicate that an ever-increasing number of people are using the Web to communicate and gather information. In 1983, there were an estimated 2,000

people using the Arpanet. In 1990, the count increased to just over 1 million users. By 2005, it was estimated that more than 900 million people worldwide would be using the Internet that year to gather and transmit information, and this figure was expected to more than double within five years. The consequence of all this growth is that decision makers now have almost instant access to large quantities of data that can be used to improve decision-making, strategic planning, and product design, and customer service. Recognizing that knowledge systems are usually based on local area network (LAN) or Internet technology, several critical questions arise when an organization attempts to implement a knowledge management system. First, how do you measure the value of a knowledge management system? Like soft-skills training, many organizations and experts are struggling to measure the value added by a knowledge management system. For example, the value of new technology in a manufacturing plant can be measured with relative accuracy and be said to decrease production costs by a certain amount per unit. Knowledge management systems, however, commonly do not have such a direct impact on operations. How can we accurately measure value of having immediate access to information that improves decision-making or strategy? Another problem is, how do you create an organizational culture that values sharing? The old adage "Information is power" exemplifies the cultural reasons why knowledge management systems can be challenging to implement. Traditionally in the United States, employees have been recognized and rewarded for individual effort and achievement. Collaborative effort and cooperation have not traditionally been rewarded. Consequently, implementing a knowledge management system may likely require that an organization reassess the values by which business is conducted, the performance evaluation instruments, and the pay/bonus structures so that employees see ample incentive to share knowledge and cooperate through-out the organization. How much information is too much? Information overload is a concern in organizations that are developing a knowledge management system. What

information do we attempt to capture and make available? What information do we overlook? In large organizations, the answers to such questions can have a dramatic impact on the quantity and quality of information available to employees. Finally, can knowledge really be captured? Knowledge managers assume that knowledge can be captured, replicated, and made useful for other members of an organization. Much knowledge, however, is tacit. It is unexpressed. For example, how do we capture the knowledge that an operations manager develops after years of working in manufacturing plants? How do we capture the sense of history, the habitual patterns of thinking, or the principles for good decision making that have proven effective over the years? If explicit knowledge is framed by tacit knowledge, how do we capture and share both forms of knowledge so that the user of the knowledge management system does not feel like the recipient of baseless or de-contextualized facts and figures?

Read more: Knowledge Management - strategy, organization, system, manager, school, company, business, system http://www.referenceforbusiness.com/management/Int-Loc/KnowledgeManagement.html#ixzz1UikARWtJ

This document is 3rd Millennium file name: Product_Literature.doc

Data Centric Knowledge Management System Product Literature

prepared by 3rd Millennium, Inc. December 4, 2003
3rd Millennium, Inc. (www.3rdmill.com) Page 2 of 6

1 Overview
The purpose of the Data Centric Knowledge Management System (KMS) is to centralize knowledge generated by scientists working within and across functional areas, and to organize that knowledge such that it can be easily accessed, searched, browsed, navigated, and curated. Rather than a complete solution, KMS is a foundation system and a framework for creating knowledge management solutions that support biopharmaceutical R&D. It is specifically designed to organize knowledge about drug targets, lead compounds, and about R&D programs. It comprises an Oracle database, and Application Programming Interface (API), a client application, and a component for automated analysis of biological sequences. Through extensions, customizations, and integration with other systems, KMS can provide useful functionality to a wide range of biopharmaceutical organizations pursuing different discovery paradigms. KMS can be used across R&D programs to: Promote controlled sharing of a companys knowledge related to biological and chemical entities. Make knowledge persistent and durable in light of organizational and personnel changes. Support the tracking of intellectual property, especially for organizations engaged in multiple collaborations. Promote better management and decision- making by providing information across projects in a consistent way in order to measure progress.
3rd Millennium, Inc. (www.3rdmill.com) Page 3 of 6

2 KMS concepts
The system is designed primarily around a few key concepts: Projects Data are organized in Projects. A project is a grouping of investigations all aimed at addressing a specific scientific question. Projects are organized in project families. Items Supported items are: Species, Strain, Chromosomal Region, Marker, Gene, Intron, Exon, 5-UTR, 3-UTR, ORF, SNP, Allele, Gene Product, Biosequence, Compound, Tissue, Population, and Phenotype. Items are also referred to as Knowledge Items. Assay Results Laboratory or computational investigations with associated results are described as assays within KMS and have one or more items associated with them. Examples of assays include Northern blots, SAGE analysis or yeast-2hybrid analyses.

3 Key Features
The system includes the following features: Browser that organizes all data available to a user in a tree structure, allowing the user to navigate through data without specifying search criteria Search interfaces for: o Assays searchable by project, assay type, creation and modification date, and user o Items searchable by project, type, assay condition, assay data, attribute, annotation, comment, external database identifier, priority, patent, sequence analysis result Item Pages that include: Summary information, Annotations, Assays, References, Patents, Related Items, and Attributes o Specialized gene page that includes Sequences and SNPs Assay Pages that include: Status, Rationale, Goal, Results, Comments, Conclusions, relationships to Knowledge Items, Attachments, Related Assays, and Assay Conditions Assay Groups that track the status of assays performed on knowledge items. Project Pages that include: Summary Information, Annotations & Attachments, and Knowledge Items. Curation interfaces for: Names, External Links, Annotations, References, Comments, Intellectual Property Analysis, and Attributes
3rd Millennium, Inc. (www.3rdmill.com) Page 4 of 6

Administration interfaces for: controlled vocabulary, templates, users, and privileges Attachments: files can be attached to annotations, assays, items, projects, and patents. Attachments can be viewed in a suitable application, if available, on the users desktop. Links to external resources: the system utilizes a flexible model for storing external database identifiers, associates these identifiers with items, and defines URLs for retrieving the content associated with the identifiers. Since the client application is written with Java/Swing, HTML pages can be rendered within the application or in an external web browser. Security model for assigning privileges to users based on types of data, projects, and project families. The access level supported are: None, Read, Write, Edit, Knowledege Reader, Knowledge Administrator, and System Administrator. Sequence Analysis Engine component: automatically processes biological sequences through bioinformatics tools, extracts relevant content from the results, and associates the results to items. The engine includes wrappers for more than 10 commonly used tools such as Blast, motif, pfam or psort and a patent search pipeline for genes. Support for assays that can be modeled as a multi-dimensional matrix relating Knowledge Items to measurements performed on the items. Genomic Map coordinates for ge nomic items. Application Programming Interface implemented in Java. Includes fine-grained and aggregate services that provide complete coverage of the database schema.

 Cross-platform Client Application: client application can be run on platforms that support Java 1.4 or greater. Includes complete source code

4 Uses of KMS

There is no single typical usage of the KMS. Rather, people with different roles are expected to leverage different features in the system. Senior scientists will make extensive use of the pages that display comprehensive summaries of information available for items, e.g., genes. The pages will present data generated by several workgroups and from the public domain, e.g., sequence data, computational biology results, SNP data, links, attachments, annotations, assay results, and references. In addition, the pages contain inferences made by researchers from the data. The pages offer a convenient point of entry for gathering the information required to investigate research questions and to formulate new research questions. Computational biologists can leverage the data and the APIs for KMS to create agents and data mining tools that derive new explicit knowledge from the complement of data and the relationships between data available in the system. For example, automated
3rd Millennium, Inc. (www.3rdmill.com) Page 5 of 6

mining of literature data may be implemented to extract references relevant to specific research interests. From these references, new knowledge may be inferred that can be loaded into KMS. Examples of such knowledge may include co-occurrences of genes with targets of interest in full-text articles or abstracts, or relationships between disease states and genes. In addition, computational biologists will utilize the assay component to record the result of the analyses they perform to identify putative targets. Program managers and senior managers can utilize KMS to track the status of projects and the advancement of work on high priority items, e.g., targets. To facilitate such work, KMS provides summary information about projects and allows recording of status information. For example, the status of assays for a subset of genes within a project is available in a project page. Furthermore, meeting minutes or summary information such as the association signal across a region may be shown in a project page.

5 Implementing and extending KMS

Implementing KMS in your organization requires careful planning. Before the system can be used productively, procedures must be defined and implemented for loading data into the system. There are two main aspects of loading data into KMS: 1) identifying data sources and relevant information within the sources; 2) defining rules that are validated when data are loaded. It is important to consider both the initial load of data to seed KMS and ongoing use. In some organizations, a curator may be responsible for processing ongoing item creation requests. For some uses, regular automated creation of items from well-defined and high-quality data sources may be appropriate. There is considerable variation in the industry of both data sources relevant to organizations, and of the rules that should be applied to the data. Therefore, KMS does not provide a standard set of loaders. Rather, the API for the system includes specific services that facilitate the creation of loaders. These services facilitate loading of assay data, items, and of data associated to items (e.g., annotations or relationships). For example, the KMS API handles and validates preferred item names, which are guaranteed unique either globally in the system or for a given project. The KMS code base includes a simple loader for LocusLink link data that can be used as an example.

KMS was designed using a modular component approach and can be extended easily beyond loaders. Examples of possible extensions include: Integration of a genome browser for visualizing the coordinates of genomic items stored in the system Integration of a small molecule viewer for visualizing and interacting with compound structures Extending the patent search pipeline to items other than genes and sequences (e.g., compounds). Creation of an alert mechanism where users are automatically alerted of new content based on interests they register with the system Adding support for new type of items (e.g., haplotype items)
3rd Millennium, Inc. (www.3rdmill.com) Page 6 of 6

Integration of a text mining tool that operates on the body of references stored in the system and creates new relationships between items Creation of a component that supports the filing of patents by extracting and formatting relevant information from the system

6 Documentation
The following documentation is provided with KMS: Product Literature (this document) System help: includes information for end users and for system administrator. The system help is provided as a stand alone document and on- line within the client application Data model and data dictionary API object models and sequence diagrams Design documents for specific components in the system Installation manual Acceptance Test Plan scenarios Licensing terms

7 Platform and system requirements

To install KMS, the following must be available: An Oracle 8i or 9i instance (see otn.oracle.com) The Ant build tool. Ant can be downloaded from ant.apache.org Java J2SE 1.4 SDK. The SDK can be downloaded from java.sun.com/j2se/downloads.html To run the KMS client, a Java Runtime Environment (JRE) version 1.4 or greater must be available. The JRE can be downloaded from (java.sun.com/j2se/downloads.html) 128 MB RAM available on the client The sequence analysis engine requires a Unix server

8 Contact information

For questions regarding the Data Centric Knowledge Management System, please contact: consulting@3rdmill.com