作者: Emitza Guzman , Omar Aly , Bernd Bruegge
DOI: 10.1109/ESEM.2015.7321214
关键词:
摘要: Context: Users can have conflicting opinions and different experiences when using software user reviews serve as a channel in which users document their experiences. To develop evolve that is usable relevant for diverse group of users, need to be taken into account. Goal: In this paper we present DIVERSE, feature sentiment centric retrieval approach automatically provides developers with sample representative the mentioned whole set reviews. Results: We evaluated diversity performance our on from seven apps two app stores. compared retrieved by DIVERSE feature-based found average outperforms baseline approach. Additionally, controlled experiment revealed help develop- ers save time analyzing was considered useful detecting evolution. Conclusions: therefore collect comprehensive aid detection opinions.