Xiaoqiao Zhang
ABSTRACT
In the Internet era, people usually consider the ratings and reviews of stores on review platforms when choosing travel locations. Today, the mainstream rating scheme for total store scores is weighted by review scores, but this scoring system can be negatively affected by malicious scoring and uneven scoring. Problems such as incompleteness and other issues will affect the authenticity of the store's rating. To this end, this paper designs a K-BERT Dianping user rating prediction based on K-BERT model to reflect real review ratings. Compared with the traditional BERT pre-training model, the K-BERT model can solve knowledge-driven problems faster through knowledge graph injection. In this paper, a Dianping knowledge map is established. Through the steps of text preprocessing, text pre-training, and Dianping dataset fine-tuning, it is found that the accuracy rate of the K-BERT model in the Dianping rating classification is about 95%. By comparing the model with BERT, Logistic Regression , it is found that the predicted effect of the K-BERT model is significantly better than the above two models.
- Yan Qiang, Meng Yue. Factors Influencing the Perceived Usefulness of Online Reviews: An Empirical Study Based on Online Film Reviews[J]. China Management Science, 2013, 21(S1): 126-131. DOI: 10.16381/j.cnki.issn1003 -207x.2013.s1.059.Google Scholar
- Qiu Xiulian, Zou Luobin, Wang Zheng.A Study on the Prediction of Film Critic Rating Based on Text Summary[J].Computer and Digital Engineering,2019,47(01):146-151.Google Scholar
- Wu Jiang, Liu Wanwan.A Study on the Influencing Factors of Online Product Reviews Based on Information Adoption Theory[J].Journal of Information Resource Management,2017,7(01):47-55.DOI:10.13365/j.jirm.2017.01.047 .Google Scholar
- ZADROZNY, SLAWOMIR, KACPRZYK, JANUSZ, GAJEWSKI, MAREK. A novel approach to sequence-of-documents focused text categorization using the concept of a degree of fuzzy set subsethood[C]. //2015 Annual Meeting of the North American Fuzzy Information Processing Society: 2015 Annual Meeting of the North American Fuzzy Information Processing Society (NAFIPS 2015), 2015 5th World Conference on Soft Computing (WConSC), August 17-19, 2015, Redmond, Washington, USA.:Institute of Electrical and Electronics Engineers, 2015:1-6.Google Scholar
- Zhang Hongli, Liu Jiying, Yang Sinan, Xu Jian. Research on Rating Prediction Model Based on Online User Reviews [J]. Data Analysis and Knowledge Discovery, 2017,1(08):48-58.Google Scholar
- Yang Guijun, Xu Xue, Zhao Fuqiang. User Rating Prediction Model and Application Based on XGBoost Algorithm [J]. Data Analysis and Knowledge Discovery, 2019,3(01):118-126.Google Scholar
- Liang Shunpan, Liu Wei, You Dianlong, Liu Zeqian, Zhang Fuzhi. A Self-Attention Capsule Network Score Prediction Model Considering Review Quality [J]. Journal of Electronics and Information, 2021,43(12):3451-3458.Google Scholar
- Zheng Jianxing, Guo Tongtong, Shen Lihua, Li Deyu.Research on recommendation method based on emotional attention of review texts[J].Journal of Zhengzhou University(Engineering Edition),2022,43(02):44-50+57.DOI:10.13705/j.issn .1671-6833.2022.02.007.Google Scholar
- YENDURI, GOKUL, RAJAKUMAR, BR, PRAGHASH, K., Heuristic-Assisted BERT for Twitter Sentiment Analysis[J]. International Journal of Computational Intelligence and Applications,2021,20(3). DOI:10.1142/S1469026821500152.Google Scholar
- Xue Yingbin. Fine-grained Short Text Sentiment Analysis Based on Bert Model [D]. Tianjin University of Technology, 2020. DOI: 10.27360/d.cnki.gtlgy.2020.000158.Google Scholar
- NING PAN, WENBIN YAO, XIAOYONG LI. Friends Recommendation Based on KBERT-CNN Text Classification Model[C]. //2021 International Joint Conference on Neural Networks: IJCNN 2021, Virtual Conference, 18-22 July 2021, [v.1].:Institute of Electrical and Electronics Engineers, 1:1-6.Devlin J , Chang M W , Lee K , BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding[J]. 2018.Google Scholar
- Fu Jing, Gong Yonggang, Lian Xiaoqin, Gu Longhao. Classification method of news short text based on BERT-LDA [J]. Information Technology and Informatization, 2021(02):127-129.Google Scholar
- Cui Jianqing, Qiu Qihao. Research on BERT Chinese News Text Classification Based on Auxiliary Tasks [J]. Software Engineering, 2022,25(06):4-8.DOI:10.19644/j.cnki.issn2096-1472.2022.006.002 .Google Scholar
- Devlin J , Chang M W , Lee K , BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding[J]. 2018.Google Scholar
- Duan Dandan, Tang Jiashan, Wen Yong, Yuan Kehai.Classification algorithm of Chinese short text based on BERT model[J].Computer Engineering,2021,47(01 ):79-86.DOI:10.19678/j.issn.1000- 3428.0056222.Google Scholar
- Li Keyue,Chen Yi,Niu Shaozhang.BERT-based social e-commerce text classification algorithm[J].Computer Science,2021,48(02):87-92.Google Scholar
- Weijie Liu,Peng Zhou,Zhe Zhao,Zhiruo Wang,Qi Ju,Haotang Deng,Ping Wang. K-BERT: Enabling Language Representation with Knowledge Graph[J]. Proceedings of the AAAI Conference on Artificial Intelligence,2020,34(03) .Google Scholar
Index Terms
- Prediction of User Ratings of Dianping Based on K-BERT Model
Recommendations
User Reviews Based Rating Prediction in Recommender System
Artificial IntelligenceAbstractThe user feedback information based Collaborative Filtering algorithm is often used to discover the hidden preferences of users and potential features of items in Recommender System for the past few years. Nevertheless, the potential relationship ...
KEP-Rec: A Knowledge Enhanced User-Item Relation Prediction Model for Personalized Recommendation
Web and Big DataAbstractFor more accurate, diversified and interpretable personalized recommendation, the joint consideration of user-item interaction information and side information in knowledge graph has become a research hotspot. Traditional models based on ...
Using inferred tag ratings to improve user-based collaborative filtering
SAC '12: Proceedings of the 27th Annual ACM Symposium on Applied ComputingUser-based collaborative filtering is one of the most widely-used recommender methods. It recommends items to a user according to her similar users' opinions. The key point of user-based collaborative filtering is to compute users' similarities. In ...
Comments