消费者决策旅程视角基于评论的深度推荐方法  

Deep review-based recommendation from the perspective of consumer decision journey

作　　者：张颖异 李先能 郭艳红[1,2] 李晓刚[1,2] 郑爽[1,2] ZHANG Ying-yi;LI Xian-neng;GUO Yan-hong;LI Xiao-gang;ZHENG Shuang(School of Economics and Management,Dalian University of Technology,Dalian 116024,China;Institute for Advanced Intelligence,Dalian University of Technology,Dalian 116024,China)

机构地区：[1]大连理工大学经济管理学院,大连116024 [2]大连理工大学先进智能研究院,大连116024

出　　处：《管理科学学报》2024年第8期105-125,共21页Journal of Management Sciences in China

基　　金：国家自然科学基金资助项目(72071029,71974031,72231010)。

摘　　要：推荐系统的本质是建模消费者行为中隐藏的兴趣偏好.人的行为离不开心理,表象的行为背后蕴含着丰富的内在产生动机.然而,当前的推荐系统主要聚焦于对消费者表象的行为数据建模层面,很少涉及内在的心理活动和决策中对信息的加工过程,难以解析行为背后的产生路径.为此,本文研究了一种新的推荐系统方法创新思路,聚焦于基于评论的推荐场景,从消费者决策旅程视角出发,引入AIDMA阶段式决策模型,将消费者行为的产生过程视为对商品“引起注意→产生兴趣→唤醒欲望→留下记忆→决策行为”的决策旅程,并将其运用到深度学习的模型设计中,提出一种新的推荐方法.实验表明本研究方法的推荐效果显著优于现有方法.本研究遵循了“模型驱动+数据驱动”的大数据驱动研究范式,不仅实现了对推荐系统具有消费者行为理论模型支撑的深度学习方法创新,还为理解消费者在决策过程中对非结构化评论文本的信息加工过程提供了借鉴和参考.The essence of recommender systems is to model the implicit preferences in consumer behavior.The human behavior is inseparable from psychology,and there are rich internal motives behind the superficial behavior.However,the current studies mainly focus on the behavioral data modeling,rarely involving the internal psychological activities and the information processing process of decision-making.Therefore,this paper studied a new idea of recommender systems by introducing AIDMA decision model from the perspective of consumer decision journey.A new deep review-based recommender system is proposed,which applies the AIDMA decision journey into the deep learning framework.Experiments showed that the recommendation performance of the proposal is significantly better than the state-of-the-art methods.This paper follows the big data-driven research paradigm of“model driven+data-driven”,realizing in-depth methodological innovation with theoretical support.

关 键 词：推荐系统 深度学习 评论文本 消费者决策旅程 

分 类 号：C93[经济管理—管理学]