Identification Method of Product Quality Problems Based on Two-view Semi-supervised Learning

doi:10.3969/j.issn.1007-7375.2023.03.010

Abstract

Abstract: Based on the abundant unstructured and unlabeled texts of consumer reviews in e-commerce websites, a two-view semi-supervised learning method is proposed to classify the reviews and identify the content related to product quality problems, so as to mine the hidden quality defects and dangers of products. Comprehensively considering the characteristics of vocabulary, emotion, domain and so on, the text view and non-text view are constructed, and the Co-training collaborative training algorithm is adopted to classify the reviews according to whether quality problems are involved. Taking the electric kettle as an example, the consumer reviews were crawled from an e-commerce website for empirical analysis. Results show that F₁ score and AUC of the proposed method are 82.18% and 86.24%, respectively, which is significantly improved compared with the single view supervised learning classifier.

Key words: reviews classification, multi-view learning, semi-supervised learning, collaborative training, quality problems identification

CLC Number:

TP393

YAO Chi, PAN Ershun. Identification Method of Product Quality Problems Based on Two-view Semi-supervised Learning[J]. Industrial Engineering Journal, 2023, 26(3): 86-94.

References

[1] KURUZOVICH J, VISWANATHAN S, AGARWAL R, et a1. Marketspace or marketplace? Online information search and channel outcomes in auto retailing[J]. Information Systems Research, 2008, 19(2): 182-201
[2] LO Y W, POTDAR V. A review of opinion mining and sentiment classification framework in social networks[C/OL]// 2009 3rd IEEE International Conference on Digital Ecosystems and Technologies. New York: IEEE, 2009 (2009-10-02). https://ieeexplore.ieee.org/abstract/document/5276705.
[3] LIU B. Sentiment analysis and subjectivity[M]//INDURKHYA N, DAMERAU F J. Handbook of Natural Language Processing. 2^nd Edition. Boca Raton: Chapman and Hall/CRC, 2010: 627-666.
[4] ZHANG W, XU H, WAN W. Weakness finder: find product weakness from Chinese reviews by using aspects based sentiment analysis[J]. Expert Systems with Applications, 2012, 39(11): 10283-10291
[5] LI Y M, CHEN H M, LIOU J H, et al. Creating social intelligence for product portfolio design[J]. Decision Support Systems, 2014, 66: 123-134
[6] LOUGHRAN T, MCDONALD B. When is a liability not a liability? Textual analysis, dictionaries, and 10-Ks[J]. The Journal of Finance, 2011, 66(1): 35-65
[7] ABBASI A, CHEN H, NUNAMAKER J F. Stylometric identification in electronic markets: scalability and robustness[J]. Journal of Management Information Systems, 2008, 25(1): 49-78
[8] ABRAHAMS A S, FAN W, WANG G A, et al. An integrated text analytic framework for product defect discovery[J]. Production and Operations Management, 2015, 24(6): 975-990
[9] LAW D, GRUSS R, ABRAHAMS A S. Automated defect discovery for dishwasher appliances from online consumer reviews[J]. Expert Systems with Applications, 2017, 67: 84-94
[10] ADAMS D Z, GRUSS R, ABRAHAMS A S. Automated discovery of safety and efficacy concerns for joint & muscle pain relief treatments from online reviews[J]. International Journal of Medical Informatics, 2017, 100: 108-120
[11] ZHANG X, QIAO Z, AHUJA A, et al. Discovering product defects and solutions from online user generated contents[C/OL]// The World Wide Web Conference. New York: Association for Computing Machinery, 2019 (2019-05-13). https://dl.acm.org/doi/abs/10.1145/3308558.3313732.
[12] ZHENG L, HE Z, HE S. A novel probabilistic graphic model to detect product defects from social media data[J/OL]. Decision Support Systems, 2020, 137: (2020-10). https://doi.org/10.1016/j.dss.2020.113369.
[13] 蒋翠清, 王齐林, 刘士喜, 等. 中文社会媒体环境下半监督学习的汽车缺陷识别方法[J]. 中国管理科学, 2014, 22(S1): 677-685
JIANG Cuiqing, WANG Qilin, LIU Shixi, et al. Semi-supervised learning for automobile defect identification in the context of Chinese social media[J]. Chinese Journal Management Science, 2014, 22(S1): 677-685
[14] 张嵩, 吴剑云, 樊卫国, 等. 基于社交媒体分析的手机缺陷识别[J]. 计算机集成制造系统, 2016, 22(9): 2264-2273
ZHANG Song, WU Jianyun, FAN Weiguo, et al. Defect discovery of phones based on social media analytics[J]. Computer Integrated Manufacturing Systems, 2016, 22(9): 2264-2273
[15] JIANG C, LIU Y, DING Y, et al. Capturing helpful reviews from social media for product quality improvement: a multi-class classification approach[J]. International Journal of Production Research, 2017, 55(12): 3528-3541
[16] 刘丹, 张玉红, 闫相斌. 在线评论的产品缺陷识别方法[J]. 黑龙江科技大学学报, 2017, 27(6): 698-704
LIU Dan, ZHANG Yuhong, YAN Xiangbin. Product defect recognition method for online reviews[J]. Journal of Heilongjiang University of Science & Technology, 2017, 27(6): 698-704
[17] 王海杰, 吴琼. 基于社交媒体的汽车缺陷识别方法[J]. 工业工程, 2018, 21(6): 31-39
WANG Haijie, WU Qiong. A defect discovery method of vehicle based on social media[J]. Industrial Engineering Journal, 2018, 21(6): 31-39
[18] 王余行, 党延忠, 徐照光. 针对论坛数据特点的汽车质量问题挖掘[J]. 中国管理科学, 2021, 29(9): 201-212
WANG Yuhang, DANG Yanzhong, XU Zhaoguang. Mining automobile quality problems based on the characteristics of forum data[J]. Chinese Journal of Management Science, 2021, 29(9): 201-212
[19] 郝媛媛, 叶强, 李一军. 基于影评数据的在线评论有用性影响因素研究[J]. 管理科学学报, 2010, 13(8): 78-88
HAO Yuanyuan, YE Qiang, LI Yijun. Research on online impact factors of customer reviews usefulness based on movie reviews data[J]. Journal of Management Sciences in China, 2010, 13(8): 78-88
[20] BLUM A, MITCHELL T. Combining labeled and unlabeled data with co-training[C/OL]// Proceedings of the eleventh annual conference on Computational learning theory. New York: Association for Computing Machinery, 1998 (1998-07-24). https://doi.org/10.1145/279943.279962.
[21] SARKAR A. Applying co-training methods to statistical parsing[C/OL]// Second Meeting of the North American Chapter of the Association for Computational Linguistics. (2001-06-02). https://dl.acm.org/doi/abs/10.3115/1073336.1073359.
[22] LIU B, FENG J, LIU M, et al. Predicting the quality of user-generated answers using co-training in community-based question answering portals[J]. Pattern Recognition Letters, 2015, 58: 29-34
[23] HONG Y, ZHU W. Spatial co-training for semi-supervised image classification[J]. Pattern Recognition Letters, 2015, 63: 59-65