用于电子商务推荐系统的混合自注意 BiLSTM 和基于激励学习的协同过滤技术

IF 3.7 4区 管理学 Q2 BUSINESS
Hemn Barzan Abdalla, Mehdi Gheisari, Ardalan Hussein Awlla
{"title":"用于电子商务推荐系统的混合自注意 BiLSTM 和基于激励学习的协同过滤技术","authors":"Hemn Barzan Abdalla, Mehdi Gheisari, Ardalan Hussein Awlla","doi":"10.1007/s10660-024-09888-5","DOIUrl":null,"url":null,"abstract":"<p>Sentiment prediction means analyzing the emotional tone or opinion contained in textual data such as reviews or posts on social media. Recommendation systems use this sentiment analysis to recommend appropriate products or content to the users. The limitations of the existing model are related to data quality and quantity issues and dealing with different scenes and languages. Hence, to overcome all these challenges, the Hybrid Self-Attention Layer Optimized Incentive Learning–based Collaborative Filtering-BiLSTM (Hybrid AT-IN based CF-BiLSTM) model was developed for sentiment prediction and recommendation based on e-commerce platforms. The usage of CF, BiLSTM networks, and a hybrid self-attention mechanism ensure the model's unrivaled precision and performance in the domain of sentiment analysis and consumer preferences. Relying on the CF, the model accumulates valuable data about user-item interactions, and BiLSTM networks process the text, employing information from the surrounding context. The model utilizes a hybrid self-attention mechanism that automatically assigns weights based on the importance of words in user reviews; this allows it to focus on the main features and improve understanding of sentiments. Moreover, applying incentive learning lets the model adapt and optimize recommendations based on changing user behaviors, leading to greater user satisfaction and engagement. In particular, the CCO-TLI model showcases significantly superior values with 98.03% accuracy, the lowest mean square error value of 1.42, 98.68% precision, 97.04% recall, and the smallest root mean squared error of 1.19 compared to existing models.</p>","PeriodicalId":47264,"journal":{"name":"Electronic Commerce Research","volume":null,"pages":null},"PeriodicalIF":3.7000,"publicationDate":"2024-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Hybrid self-attention BiLSTM and incentive learning-based collaborative filtering for e-commerce recommendation systems\",\"authors\":\"Hemn Barzan Abdalla, Mehdi Gheisari, Ardalan Hussein Awlla\",\"doi\":\"10.1007/s10660-024-09888-5\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>Sentiment prediction means analyzing the emotional tone or opinion contained in textual data such as reviews or posts on social media. Recommendation systems use this sentiment analysis to recommend appropriate products or content to the users. The limitations of the existing model are related to data quality and quantity issues and dealing with different scenes and languages. Hence, to overcome all these challenges, the Hybrid Self-Attention Layer Optimized Incentive Learning–based Collaborative Filtering-BiLSTM (Hybrid AT-IN based CF-BiLSTM) model was developed for sentiment prediction and recommendation based on e-commerce platforms. The usage of CF, BiLSTM networks, and a hybrid self-attention mechanism ensure the model's unrivaled precision and performance in the domain of sentiment analysis and consumer preferences. Relying on the CF, the model accumulates valuable data about user-item interactions, and BiLSTM networks process the text, employing information from the surrounding context. The model utilizes a hybrid self-attention mechanism that automatically assigns weights based on the importance of words in user reviews; this allows it to focus on the main features and improve understanding of sentiments. Moreover, applying incentive learning lets the model adapt and optimize recommendations based on changing user behaviors, leading to greater user satisfaction and engagement. In particular, the CCO-TLI model showcases significantly superior values with 98.03% accuracy, the lowest mean square error value of 1.42, 98.68% precision, 97.04% recall, and the smallest root mean squared error of 1.19 compared to existing models.</p>\",\"PeriodicalId\":47264,\"journal\":{\"name\":\"Electronic Commerce Research\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":3.7000,\"publicationDate\":\"2024-08-20\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Electronic Commerce Research\",\"FirstCategoryId\":\"91\",\"ListUrlMain\":\"https://doi.org/10.1007/s10660-024-09888-5\",\"RegionNum\":4,\"RegionCategory\":\"管理学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"BUSINESS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Electronic Commerce Research","FirstCategoryId":"91","ListUrlMain":"https://doi.org/10.1007/s10660-024-09888-5","RegionNum":4,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"BUSINESS","Score":null,"Total":0}
引用次数: 0

摘要

情感预测是指分析社交媒体上的评论或帖子等文本数据中包含的情感基调或观点。推荐系统利用这种情感分析向用户推荐合适的产品或内容。现有模型的局限性与数据质量和数量问题以及处理不同场景和语言有关。因此,为了克服所有这些挑战,我们开发了基于自关注层优化激励学习的混合协同过滤-BiLSTM(Hybrid AT-IN based CF-BiLSTM)模型,用于基于电子商务平台的情感预测和推荐。通过使用 CF、BiLSTM 网络和混合自注意机制,确保了该模型在情感分析和消费者偏好领域具有无与伦比的精度和性能。依靠 CF,该模型积累了关于用户与商品互动的宝贵数据,而 BiLSTM 网络则利用周围上下文的信息处理文本。该模型采用了一种混合自我关注机制,可根据用户评论中词语的重要性自动分配权重;这使其能够专注于主要特征并提高对情感的理解。此外,激励学习的应用使模型能够根据用户行为的变化调整和优化推荐,从而提高用户满意度和参与度。与现有模型相比,CCO-TLI 模型的准确率为 98.03%,均方误差最小值为 1.42,精确率为 98.68%,召回率为 97.04%,均方根误差最小值为 1.19。
本文章由计算机程序翻译,如有差异,请以英文原文为准。

Hybrid self-attention BiLSTM and incentive learning-based collaborative filtering for e-commerce recommendation systems

Hybrid self-attention BiLSTM and incentive learning-based collaborative filtering for e-commerce recommendation systems

Sentiment prediction means analyzing the emotional tone or opinion contained in textual data such as reviews or posts on social media. Recommendation systems use this sentiment analysis to recommend appropriate products or content to the users. The limitations of the existing model are related to data quality and quantity issues and dealing with different scenes and languages. Hence, to overcome all these challenges, the Hybrid Self-Attention Layer Optimized Incentive Learning–based Collaborative Filtering-BiLSTM (Hybrid AT-IN based CF-BiLSTM) model was developed for sentiment prediction and recommendation based on e-commerce platforms. The usage of CF, BiLSTM networks, and a hybrid self-attention mechanism ensure the model's unrivaled precision and performance in the domain of sentiment analysis and consumer preferences. Relying on the CF, the model accumulates valuable data about user-item interactions, and BiLSTM networks process the text, employing information from the surrounding context. The model utilizes a hybrid self-attention mechanism that automatically assigns weights based on the importance of words in user reviews; this allows it to focus on the main features and improve understanding of sentiments. Moreover, applying incentive learning lets the model adapt and optimize recommendations based on changing user behaviors, leading to greater user satisfaction and engagement. In particular, the CCO-TLI model showcases significantly superior values with 98.03% accuracy, the lowest mean square error value of 1.42, 98.68% precision, 97.04% recall, and the smallest root mean squared error of 1.19 compared to existing models.

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
CiteScore
7.50
自引率
12.80%
发文量
99
期刊介绍: The Internet and the World Wide Web have brought a fundamental change in the way that individuals access data, information and services. Individuals have access to vast amounts of data, to experts and services that are not limited in time or space. This has forced business to change the way in which they conduct their commercial transactions with their end customers and with other businesses, resulting in the development of a global market through the Internet. The emergence of the Internet and electronic commerce raises many new research issues. The Electronic Commerce Research journal will serve as a forum for stimulating and disseminating research into all facets of electronic commerce - from research into core enabling technologies to work on assessing and understanding the implications of these technologies on societies, economies, businesses and individuals. The journal concentrates on theoretical as well as empirical research that leads to better understanding of electronic commerce and its implications. Topics covered by the journal include, but are not restricted to the following subjects as they relate to the Internet and electronic commerce: Dissemination of services through the Internet;Intelligent agents technologies and their impact;The global impact of electronic commerce;The economics of electronic commerce;Fraud reduction on the Internet;Mobile electronic commerce;Virtual electronic commerce systems;Application of computer and communication technologies to electronic commerce;Electronic market mechanisms and their impact;Auctioning over the Internet;Business models of Internet based companies;Service creation and provisioning;The job market created by the Internet and electronic commerce;Security, privacy, authorization and authentication of users and transactions on the Internet;Electronic data interc hange over the Internet;Electronic payment systems and electronic funds transfer;The impact of electronic commerce on organizational structures and processes;Supply chain management through the Internet;Marketing on the Internet;User adaptive advertisement;Standards in electronic commerce and their analysis;Metrics, measurement and prediction of user activity;On-line stock markets and financial trading;User devices for accessing the Internet and conducting electronic transactions;Efficient search techniques and engines on the WWW;Web based languages (e.g., HTML, XML, VRML, Java);Multimedia storage and distribution;Internet;Collaborative learning, gaming and work;Presentation page design techniques and tools;Virtual reality on the net and 3D visualization;Browsers and user interfaces;Web site management techniques and tools;Managing middleware to support electronic commerce;Web based education, and training;Electronic journals and publishing on the Internet;Legal issues, taxation and property rights;Modeling and design of networks to support Internet applications;Modeling, design and sizing of web site servers;Reliability of intensive on-line applications;Pervasive devices and pervasive computing in electronic commerce;Workflow for electronic commerce applications;Coordination technologies for electronic commerce;Personalization and mass customization technologies;Marketing and customer relationship management in electronic commerce;Service creation and provisioning. Audience: Academics and professionals involved in electronic commerce research and the application and use of the Internet. Managers, consultants, decision-makers and developers who value the use of electronic com merce research results. Special Issues: Electronic Commerce Research publishes from time to time a special issue of the devoted to a single subject area. If interested in serving as a guest editor for a special issue, please contact the Editor-in-Chief J. Christopher Westland at westland@uic.edu with a proposal for the special issue. Officially cited as: Electron Commer Res
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信