Use this URL to cite or link to this record in EThOS:
Title: Text mining for meeting transcripts analysis to support decision management
Author: Chibelushi, Caroline
ISNI:       0000 0001 2426 4009
Awarding Body: Staffordshire University
Current Institution: Staffordshire University
Date of Award: 2008
Availability of Full Text:
Access from EThOS:
This thesis presents research that has developed a methodology for extracting the elements of decision making from transcripts generated from software development project meetings. These elements, namely: agents, needs, issues aI1d actions are necessary in supporting decision management systems. This research considers decision management systems as a means to control rework in system developments projects. Rework involves altering, revising or a complete re-doing ofa certain project activity because the previous work was incorrect, incomplete or inconsistent. It is a major cost factor in system development projects that accounts for. over 50% of additional efforts and substantial costs, particularly for large projects. Rework can be associated with incorrect or inappropriate decision making. However, decisions could be 'hidden' and difficult to identify, and the evidence of their existence is related to the issues discussed, the meeting participants' needs, the actions taken to satisfy the participants' needs, as well as the agents who will perform the chosen actions. Such detailed information is often not found in most minutes of meetings, and may result in unmanaged decisions. This research argues that by investigating the recorded and transcribed meeting conversations, it is possible to identify and extract these evidences from meeting decision making processes, referred to herein as the elements of decision making. By understanding the relationship between these elements, managers will be able to detect any incorrect decisions and communication failures, and to understand their impact on rework. Methods which are associated with the decision making process are developed in this thesis; which are then used to extract the above-mentioned elements. The thesis introduces the concept of decision making elements, which are then incorporated into the decision model entitled ANIA (Agents, Needs, Issues and Actions) and employed in a text mining methodology. The text mining methodology used in this thesis combines language understanding levels (syntax, semantics, discourse analysis and aspects of pragmatics) with statistical methods (vector space model and word frequency occurrences) to identify and extract these four elements. The methodology is validated using seven transcripts of meetings related to software development projects. The evaluation results show that the methodology is able to detect the topics discussed in the meetings and then to extract the four elements of decision making.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available