A Scenario Model Aggregation Approach for Mobile App Requirements Evolution Based on User Comments

Abstract

With the increasingly intense competition in mobile applications, more and more attention has been paid to online comments. For the masses, comments have been viewed as reliable references to guide the choice of applications; for providers, they have been regarded as an important channel to learn expectations, demands and complaints of users. Therefore, comments analysis has become a hot topic in both requirements engineering and mobile application development. But analyzers in both areas are always not only suffered from the vast noise in comments, but also troubled by their incompleteness and inaccuracy. Therefore, how to obtain more convincing enlightenments from comments and how to reduce the manpower needed become the research focuses. This paper aims to propose a Scenario Model Aggregation Approach (SMAA) for analyzing and modeling user comments of mobile applications. By selecting appropriate natural language processing technologies and machine learning algorithms, SMAA can help requirements analysts to build aggregated scenario models, which can be used as the source of evolutionary requirements for the decision making of application evolution. The aggregated scenario model is not only easy to read and understand, but also able to reduce the manpower needed greatly. Finally, the feasibility of SMAA is exemplified by a case study.