Advanced search
Start date
Betweenand
Related content
(Reference retrieved automatically from Web of Science through information on FAPESP grant and its corresponding number as mentioned in the publication by the authors.)

Opinion mining for app reviews: an analysis of textual representation and predictive models

Full text
Author(s):
Araujo, Adailton F. [1] ; Golo, Marcos P. S. [1] ; Marcacini, Ricardo M. [1]
Total Authors: 3
Affiliation:
[1] Univ Sao Paulo, Inst Math & Comp Sci, USP, POB 668, BR-13560970 Sao Carlos, SP - Brazil
Total Affiliations: 1
Document type: Review article
Source: AUTOMATED SOFTWARE ENGINEERING; v. 29, n. 1 MAY 2022.
Web of Science Citations: 0
Abstract

Popular mobile applications receive millions of user reviews. These reviews contain relevant information for software maintenance, such as bug reports and improvement suggestions. The review's information is a valuable knowledge source for software requirements engineering since the apps review analysis helps make strategic decisions to improve the app quality. However, due to the large volume of texts, the manual extraction of the relevant information is an impracticable task. Opinion mining is the field of study for analyzing people's sentiments and emotions through opinions expressed on the web, such as social networks, forums, and community platforms for products and services recommendation. In this paper, we investigate opinion mining for app reviews. In particular, we compare textual representation techniques for classification, sentiment analysis, and utility prediction from app reviews. We discuss and evaluate different techniques for the textual representation of reviews, from traditional Bag-of-Words (BoW) to the most recent state-of-the-art Neural Language models (NLM). Our findings show that the traditional Bag-of-Words model, combined with a careful analysis of text pre-processing techniques, is still competitive. It obtains results close to the NLM in the classification, sentiment analysis and utility prediction tasks. However, NLM proved to be more advantageous since they achieved very competitive performance in all the predictive tasks covered in this work, provide significant dimensionality reduction, and deals more adequately with semantic proximity between the reviews' texts. {[}GRAPHICS] . (AU)

FAPESP's process: 19/25010-5 - Semantically enriched representations for Portuguese textmining: models and applications
Grantee:Solange Oliveira Rezende
Support Opportunities: Regular Research Grants
FAPESP's process: 19/07665-4 - Center for Artificial Intelligence
Grantee:Fabio Gagliardi Cozman
Support Opportunities: Research Grants - Research Program in eScience and Data Science - Research Centers in Engineering Program