Feedback Recommendation System Based on Structured Feedback Acquisition

Abstract

Users' feedback are increasingly becoming a vital factor during the software evolution process. Users' feedback are widely considered to be an indication of how satisfied are the system's users. It was found that large percentage of feedback represents problems reported by users who face problems in using the system. Users' feedback are usually communicated in an ad-hoc manner with no defined structure. The ad-hoc nature of the feedback makes it difficult to study and analyse feedback. Also, it usually happens that similar problems are reported by different users, so engineers spend wasted time and effort in duplicate issues. In this paper we propose a recommendation system which recommends to the user reported problems that are similar to his from prevoius feedback threads to reuse existing solutions. We based our work on a structured feedback system to ensure we get better results. Structured feedback ensures minimum level of useful and meanigful information that helps enhance the analysis results. Initial evaluation was conducted and the results are promising.