Evaluating a Software Engineering Knowledge Base

Show full item record

Title: Evaluating a Software Engineering Knowledge Base
Author: Baheti, Prashant P
Advisors: Dr. Ana I. Anton, Committee Member
Dr. Aldo Dagnino, Committee Member
Dr. Laurie Williams, Committee Chair
Abstract: Often, a large amount of knowledge leaves the organization when an employee and the knowledge he or she possesses leave the organization. Knowledge Management (KM) facilitates the retention and distribution of intellectual capital within an organization and helps the organization to gain competitive advantage and create business value. Organizational learning occurs when knowledge from all projects is documented and stored in a repository to aid future decision making. Such an information repository is called a Knowledge Base (KB). A KB is an important element of any KM system. This thesis has the following objectives: 1. to describe the re-engineering of a software engineering KB within an organization; 2. to identify the critical elements that influence the success of a software engineering KB within an organization based on a case study at ABB; and 3. to propose a metric suite for evaluating a KB and to partially validate this metric suite using a case study. This thesis proposes a set of metrics for evaluating a software engineering KB. The main goals of evaluating a KB are to determine whether the presence of the KB has benefited the organization (ABB) and to identify areas for improvement of the KB. The KB evaluation can be seen in the context of the Goal/Question/Metric (GQM) model. Via GQM, we set goals, refining the goals to a set of questions, and answer these questions through a set of metrics. ABB had an existing Experience Database (ABB-1) containing experiences submitted by the employees as well as training materials. However, ABB-1 was not used by the employees as intended because of certain deficiencies. As part of this research, these deficiencies were identified within the context of the KM reference model proposed by Abou-Zeid. This information was useful in the re-engineering of the ABB-1 to the ABB Software Process Initiative KB (ABB-2). The Quality Improvement Paradigm guided the re-engineering process. The metrics collected indicate that ABB-2 is being presently used by only a small number of users. Although four out of five users felt that ABB-2 made it easy to find useful information, all the five users believe that ABB-2 produces useful information less than 60% of the times. However, the projected timesaving because of presence of ABB-2 are much greater than the time invested in developing and evaluating ABB-2, even by taking conservative estimates. In our research, we have found four factors that are instrumental to the success of a KB. 1. assigning formal responsibility of the maintaining the KB to one or more persons; 2. proper publicity of the KB within the organization; 3. ability to easily find resources on KB, making it conducive to visit the KB for knowledge needs; and 4. employee participation in contributing to and regularly updating the KB. The ABB case study can serve as a guide to implementing a KB within an organization. Partial validation has been performed for a subset of the proposed metrics. Since data collection mechanisms are still in place, a complete validation of the metrics will be completed after ABB-2 is regularly used by large number of employees.
Date: 2004-08-15
Degree: MS
Discipline: Computer Science
URI: http://www.lib.ncsu.edu/resolver/1840.16/499


Files in this item

Files Size Format View
etd.pdf 734.0Kb PDF View/Open

This item appears in the following Collection(s)

Show full item record