[1]汪 烨,周思源,翁知远,等.一种面向用户反馈的智能分析与服务设计方法[J].郑州大学学报(工学版),2023,44(03):58-63.[doi:10.13705/j.issn.1671-6833.2022.06.004]
　Wang Ye,Zhou Siyuan,Weng Zhiyuan,et al.An Intelligent Analysis and Service Design Method for User Feedback[J].Journal of Zhengzhou University (Engineering Science),2023,44(03):58-63.[doi:10.13705/j.issn.1671-6833.2022.06.004]
点击复制

一种面向用户反馈的智能分析与服务设计方法()
分享到：

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:
44
期数:
2023年03期
页码:
58-63
栏目:
出版日期:
2023-04-30

文章信息/Info

Title:
An Intelligent Analysis and Service Design Method for User Feedback
作者:
汪 烨1;  周思源1;  翁知远2;  陈骏武1
1．浙江工商大学 计算机与信息工程学院，浙江 杭州 310018; 2．内布拉斯加大学林肯分校 计算机科学与工程系，美 国内布拉斯加州 林肯市 68508

Author(s):
Wang Ye1;  Zhou Siyuan1;  Weng Zhiyuan2;  Chen Junwu1
1.School of Computer and Information Engineering, Zhejiang University of Technology, Hangzhou 310018, Zhejiang, 2.Department of Computer Science and Engineering, University of Nebhauska, Lincoln City, Inner Branch

关键词:
需求分析;  服务设计;  服务计算;  深度学习;  用户反馈
分类号:
TP311
DOI:
10.13705/j.issn.1671-6833.2022.06.004
文献标志码:
A
摘要:
针对用户评论数据，提出了一种面向用户反馈的智能分析与服务设计方法。该方法选取了 IOS 平台多个 App 的用户评论数据，对其进行智能挖掘和分类，分析其中的潜在需求。首先，分析用户需求类别，将划分的 10 个 需求进行具体定义。其次，对用户数据进行爬取、清洗和标注，形成软件分类数据集。通过实验检验 TextCNN、BiLSTM_Attention 和 BEＲT 对用户评论数据智能分类的效果，将分类结果进行优先级排序。最后，将该方法封装成一 种可重用的智能服务供使用者远程调用。实验结果表明: TextCNN 模型综合效果最好，在单一指标 Precision 上， BEＲT 模型效果最好; BEＲT 模型利用并行计算优化训练过程，使其可拓展到大规模项目，在数据量大、精确性要求 比较高的情况下，推荐 BEＲT 模型; 反之，在应对数据小、时限紧的情况时，推荐 TextCNN 模型。
Abstract:
In recent years, intelligent services in various fields have been paid more and more attention and achieved rapid development. There are a lot of user comment data in the use feedback of each application software. It is of great significance to mine meaningful user demands from these large numbers of user comment data with varying quality. At present, the existing requirements intelligent classification methods do not improve the reusability and application value from the perspective of service computing. Therefore, it is an important direction to study a method of intelligent mining of user requirements and design it as reusable service to help the iterative update of application software. This paper focuses on the design method of an intelligent service to mine user needs, selects user review data of multiple apps on IOS platform, conducts intelligent mining and classification on them, and analyzes the potential demands. First, analyze the user demand category, define the demand category Then, the user data is crawled, cleaned and annotated to form a software classification data set. Then, the effect of mainstream deep learning (TextCNN, BiLSTM_Attention and BERT) on intelligent classification of user comment data is explored through experiments, and the classification results are prioritized. Finally, the intelligent service is designed and packaged as a Python package that can be invoked. Through experimental comparison, it is found that BERT model performs better in Precision, Recall and F1-Measure.

参考文献/References:

［1］ LEE G，ＲAGHU T S． Product portfolio and mobile apps success: evidence from app store market［J］． Journal of Management Information Systems，2014，31 ( 2) : 133－170． 

［2］ HENZE N，BOLL S． Ｒelease your app on Sunday eve: finding the best time to deploy apps［C］∥Proceedings of the 13th International Conference on Human Computer Interaction with Mobile Devices and Services． New York: ACM，2011: 581－586． 
［3］ SCHNEIDEＲ K． Focusing spontaneous feedback to support system evolution［C］∥ 19th International Ｒequirements Engineering Conference． Piscataway: IEEE， 2011: 165－174．
 ［4］ CHEN N，LIN J L，HOI S C H，et al． AＲ-miner: mining informative reviews for developers from mobile app marketplace［C］∥Proceedings of the 36th International Conference on Software Engineering． New York: ACM， 2014: 767－778．
 ［5］ PANICHELLA S，SOＲBO A D，GUZMAN E，et al． How can i improve my app? Classifying user reviews for software maintenance and evolution［C］∥IEEE International Conference on Software Maintenance ＆ Evolution． Pisca-taway: IEEE，2015: 281－290． 
［6］ 王莹，郑丽伟，张禹尧，等．面向中文 APP 用户评论数据的软件需求挖掘方法［J］． 计 算 机 科 学，2020，47 ( 12) : 56－64．
 WANG Y，ZHENG L W，ZHANG Y Y，et al． Software requirement mining method for Chinese APP user review data［J］．Computer Science，2020，47( 12) : 56－64 ．
［7］ KIM Y． Convolutional neural networks for sentence classification［EB/OL］． ( 2014 － 08 － 25) ［2021 － 12 － 09］． https: ∥arxiv．org /abs/1408. 5882． 
［8］ ZHOU P，SHI W，TIAN J，et al． Attention-based bidirectional long short-term memory networks for relation classification［C］∥ Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics． Stroudsburg: Association for Computational Linguistics， 2016: 207－212．
 ［9］ VASWANI A， SHAZEEＲ N， PAＲMAＲ N， et al． Attention is all you need［EB/OL］．( 2017－12－06) ［2021－ 12－09］． https: ∥arxiv．org /abs/1706. 03762． 
［10］ 贾一荻，刘璘． 中文非功能需求描述的识别与分类方 法研究［J］． 软件学报，2019，30( 10) : 3115－3126．
 JIA Y D，LIU L． Ｒecognition and classification of nonfunctional requirements in Chinese ［J］． Journal of Software，2019，30( 10) : 3115－3126． 
［11］ KUＲTANOVIC Z，MAALEJ W． Automatically classifying functional and non-functional requirements using supervised machine learning［C］∥2017 IEEE 25th International Ｒequirements Engineering Conference． Piscataway: IEEE，2017: 490－495． 
［12］ HUANG Q，XIA X，XING Z C，et al． API method recommendation without worrying about the task-API knowledge gap［C］∥Proceedings of the 33rd ACM/IEEE International Conference on Automated Software Engineering． New York: ACM，2018: 293－304． 
［13］ ＲAHMAN M M，ＲOY C K，LO D． ＲACK: automatic API recommendation using crowdsourced knowledge［C］ ∥2016 IEEE 23rd International Conference on Software Analysis， Evolution， and Ｒeengineering． Piscataway: IEEE，2016: 349－359.

更新日期/Last Update: 2023-05-09