Abstract

User feedback is mainly defined as an information source for evaluating the customers’ satisfaction for a given goods, service or software application. Due to the wide diffusion of the Internet and to the proliferation of mobile devices, users access a myriad of software services and applications, at any time and in any place. In this context users can provide feedback upon their experience in using software, through dedicated software applications or web forms. This online user feedback is a powerful source of information for improving the software service or application. Specifically in software engineering, user feedback is recognized as a source of requests for change in a system, so it can contribute to the evolution of software systems. Indeed, user feedback is gaining more attention from the requirements engineering research community, and dedicated buzzwords have been introduced to refer to research studies in RE, i.e. mass RE and crowd RE. Arguing on this premise, the possibility of exploiting user feedback is worth to be investigated in requirements engineering, by addressing open challenges in collection as well as in the analysis of online feedback. The research work described in this Thesis starts with a stateof-the-art literature analysis that revealed that the definition of user feedback as an artifact, as well as the characterization and understanding of its process of elaboration and communication were still unexplored, especially from the requirements engineering perspective. We adopted a multidisciplinary approach by borrowing concepts and techniques from ontologies, philosophy of language, natural language processing, requirements engineering and human computer interaction. The main research contributions are: an ontology of user feedback, the characterization of user feedback as speech acts for applying a semantic analysis, and the proposal of a new way of gathering and filtering user feedback by applying an argumentation framework.