摘要
面对数量巨大、信息非结构化的App用户评论,获取软件特征有助于App开发者进一步探索用户的特征请求及潜在需求。文章针对App用户评论的特性,提出新的App软件特征挖掘方法,以Apple App Store里真实的App用户评论为实验对象,收集、预处理评论数据,并进行特征挖掘。实验结果显示,识别特征请求评论范围具有一定的必要性以及提取软件特征具有较高的准确性,由此验证了本文方法的有效性。
In accordance with the characteristics of App user reviews, the paper proposes a new App software feature mining method. Taking the real App user reviews in Apple App Store as experiment subjects, the paper collects and preprocesses a large number of reviews, then uses the above method to mine the software features. The experimental results show that it is necessary to identify the comment range of feature request and it has a high accuracy of the extraction of software features, which verify the effectiveness of the proposed method.
作者
吕宏玉
樊坤
杨建林
Lv Hong-yu;Fan Kun;Yang Jian-lin
出处
《图书馆理论与实践》
CSSCI
2019年第7期106-112,共7页
Library Theory and Practice
基金
2018年系国家社会科学基金重点项目“面向国家发展与安全决策的情报服务创新研究”(项目编号:18ATQ003)的研究成果
