Evaluation of Project Work Using Clustering and Pattern Mining of Online Collaborative Data

Abstract

Group work is widespread in software development life cycle. The huge data can be generated by increasing use of online tools which supports group work. The aim is to exploit this group data to evaluate performance with the presentation of constructive high-level views of information about the team, along with desirable patterns identifying the behavior of strong groups. The key purpose of this work is to facilitate the groups and their coordinators, so that they can observe the applicable aspects of the software team’s functions, provide feedback and point out where are the issues. The context for this work is software development project where team members use online collaborative tools. The clustering and pattern mining techniques can be useful to explore the possibilities of finding the patterns distinguishing strong group from weaker group and get the vision of the factors such as importance of leadership and effective team interaction, which could lead to the success of group.  The results can be helpful to offer recommendations for potential and poor practices at the early stages accompanied with remediation of poor practices.