A Research Review of Mobile Application Review Mining

[Purpose/significance] User reviews are helpful for developers to realize mobile application innovation. This paper summarizes the literature related to mobile application review mining and provides references for mobile application development and review mining. [Method/process] This study reviewed...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhang Ji, Kang Lele, Li Bo
Format: article
Language:ZH
Published: LIS Press 2021
Subjects:
Z
Online Access:https://doaj.org/article/6866f07e8a1f403dbf395c1e3bfd300f
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:[Purpose/significance] User reviews are helpful for developers to realize mobile application innovation. This paper summarizes the literature related to mobile application review mining and provides references for mobile application development and review mining. [Method/process] This study reviewed the researches related to mobile application review mining into three key themes of review classification, review clustering and review feature extraction by using the text analysis method, and expounded on the development status of this field according to this framework. [Result/conclusion] At present, the methods of review classification have begun to evolve from machine learning to deep learning; review clustering mainly uses K-Means and DBSCAN; feature extraction is still focused on the explicit features of APP reviews. In the future, there are still three issues worth exploring in mobile application review mining: domain dependence, multi-source information fusion and review value evaluation.