You are on page 1of 7

Business Requirements Document (BRD)

Version V 1

Data Analysis-Plan Module (Using the XML)

Course5 Intelligence
1 Document Revisions

Version
Date Document Changes
Number
11/02/2020 V1 Version V1

2 Introduction

Analysis Plan need to be built as per the XML, for this all the functionality and special instruction will be
same as the previous version (with Q’re) of the analysis plan. In this module input file will be XML file
(With or without Q’re) and output will be same i.e. analysis plan builder with auto detect of
question types detection and special instruction.
It includes the client standards and custom analysis requirement. Few are as follows,
1. Table Number
2. Table Title
3. Base Description and condition
4. Nets (i.e. grouping of responses)
5. Statistics (Mean, Median, STD Err, STD Dev etc.).
6. Summary tables requirements
7. Customs tables requirements
8. Grid Table Requirements (2D and 3D)
9. Weighting Specifications
10. Banner OR Cross-break definitions
11. Significance test requirements.
12. Deliverables and its format requirements

2.1 Purpose
The purpose of this document is to discuss complete functionality of the data analysis module using
xml as the base.

2.2 Scope
The DI OS from Course5 Intelligence is designed for officialdoms who are looking to bridge the data
tables and analysis plan robotically by means of Artificial Intelligence. There will be different Module
will be define the functionality of the data analysis module.
In this module, the scope will be elaborated and we have to consider the XML as the input file.
2.3 Objective
With AI automation, using the survey input files i.e. XML and Data files, it should produce a draft
version of Analysis Plan based on type of question for each question in the survey. This will
benefit the client services team to reduce any manual intervention to design the analytics plan for
each question of surveys and results into saving of their efforts.
Every client has their own standards on format of tables and their requirements, which will be
controlled via client settings.

2.4 DI Flow: - flow will be same as the previous version.


Data Analysis

Input Materials for the Data Analysis Module

A. Survey XML

Extensible Markup Language (XML) is a markup language that defines a set of rules for encoding
documents in a format that is both human-readable and machine-readable. It has finalized survey
schema with logics and definitions. This is an optional input file, as we may not be getting survey
xml for all the projects from the clients.
Our DI-OS must need to read all different SP tools xml file like confirmit, Decipher, Askia, Qualtrics,
Dimensions etc., but to begin we are focusing on confirmit and Decipher and rest will be focused in
the next phase.
B. Data File
It has respondent level data collected and data structure
Our DI-OS must need to read all different format of data files (SPSS, Excel, ASCII, Askia and
dimension data files). To begin with, we are focusing on SPSS Decipher and rest will be focused in
the next phase.
i) Variable Name and Question Text
ii) Question Responses answer (with response codes)
iii) Only Dimension data files has survey structure in it
iv) For ASCII data file it should read data file and data map.
v) Data Variable to use for Table view
vi) Width and decimal point of the variables
vii) For excel data file it should read both code and label
viii) For Askia data file it should read the .qes

Working Mechanism for Analysis Plan


Step 1 - AI should identify the Question #, Question Text
Step 2 - AI should identify the Question type of Question using XML and Data structure
Step 3 - AI should validate OR match Question # of survey with Variable names of SPSS data, to pick
the answer options and its response codes of the respective questions.
Step 4 - AI should read the question logic to define the table base descriptions and logic from the
Q’re if it is provided.
Step 5 – Based on Question Type and client standards (which will be defined later), it should
allocate required analysis requirements of survey.

It should auto populate


a) Question # and Question Text
b) Question skipping condition OR Base condition (from Q’re if available)
c) Nets and statistics requirements
d) Formatting requirements (if any) like ranking or sorting and
e) Client standard requirement based on client standards and Question type (which will be
defined later).

Rules for identifying the Question type: - identification will also be same as per the
previous module
Special instruction for the each question type will also be same as per the previous
module.

Analysis Plan UI- Standard Tables

UI will be same as per the developed version; difference between the modules will be as
follows.
1) Question ID: - will be generated based on the question ID in the XML. System will
read all the tags from the xml and according to the nametag, the id of the question
will be decided.

For reference:-
<variable ident="115" type="single">
<name>S1</name>
<label>S1 - Do you work in any of the following occupations?</label>
<position start="8453" finish="8454" />
<values>
<value code="1">Advertising</value>
<value code="2">Market Research/Marketing</value>
<value code="3">Journalism / Public Relations</value>
<value code="4">Commercial Estate Agency</value>
<value code="5">Shopping Centre management</value>
<value code="6">None of these</value>
<value code="99">Do not wish to answer</value>
</values>
</variable>

In the above reference XML tag the nametag “<name>S1</name>”, so here


question ID will be “S1”.

2) Question type: - in the XML, the question type will be generated according to type
tag. For example: -<variable ident="115" type="single">, in the reference
example the type of the question will be “single”.
Note: - the type tag will always be enclosed with the “variable ident”. It is nothing
but a unique number created by the XML for each variable.

3) Table title: - it will be generated from the tag form label. For reference “<label>S1 -
Do you work in any of the following occupations? </label>”, so in the referred
example the table tittle will be “S1 - Do you work in any of the following
occupations?”

4) Value labels: - the value label and value code will be in the value tag in the XML file.
For reference: - <value code="1">Advertising</value>
<value code="2">Market Research/Marketing</value>
<value code="3">Journalism / Public Relations</value>
<value code="4">Commercial Estate Agency</value>
<value code="5">Shopping Centre management</value>
<value code="6">None of these</value>
<value code="99">Do not wish to answer</value>
So here code will be
1 for “Advertising”
2 for “Market Research/Marketing”
3 for “Journalism / Public Relations”
4 for “Commercial Estate Agency”
5 for “Shopping Centre management”
6 for “None of these”
99 for “Do not wish to answer”

5) Base Description: - If Q’re is available along with the Q’re then base description
should be picked from it and if not by it should checked the data file if there is
response for all respondents then it should populate as “All respondent”, if there is
any respondent less than the total then it should populate as “Total Answering”. In
addition, user has to change the same manually if user wants to change.

6) Base definition: - If Q’re is available along with the Q’re then base condition should
be picked from it, if not then it should define as “Not defined”. In addition, user has
to change the same manually if user wants to change.

7) Summary: - Summary will be defined all selection of the user for the generation of
the table script.
Data Comparison Module
Data comparison module will be same as per the previous version, as we are comparing the
data file, not the xml. We will decide in the next release, how we gone compare the XML.

Data Labelling Module

Data Labelling module will be same as per the previous version, labelling will be done
according to the XML. We will decide the UI for the data labelling in the next phase.
Labelling will be done based on the xml.

You might also like