杨金庆 程秀峰 周玮珽

摘 要：[目的/意义]情境感知推荐是推荐系统发展过程中的重要阶段，其核心思想是将情境信息融入资源推荐的过程中，以此提高资源推荐系统的准确性。作为资源推荐的重要技术，情境感知推荐吸引了各领域的广泛关注，逐渐成为资源推荐研究的重要组成部分。[方法/过程]本文以筛选于WOS、CNKI等数据库的文献集为研究对象，通过归纳总结的方法，首先系统梳理、总结了情境感知推荐的主要方法—情境前过滤推荐方法、情境后过滤推荐方法以及情境建模方法，然后从情境数据采集和情境感知应用两个方面总结分析情境感知推荐实践进展，最后总结情境及情境感知技术对资源推荐研究的理论价值与实践的现实意义。[结果/结论]未来研究将综合3种主流情境感知资源推荐方法，以混合形式融合情境要素，构建多维、复杂的情境推荐模型，进一步提升精细化、个性化信息资源服务体验。

关键词：情境感知;资源推荐;情境感知推荐

DOI：10.3969/j.issn.1008-0821.2020.02.017

〔中图分类号〕G250.73 〔文献标识码〕A 〔文章编号〕1008-0821（2020）02-0153-07

Research Review and Progress on Practice of the Resource

Recommendation Research Based on Context-awareness

Yang Jinqing1 Cheng Xiufeng2 Zhou Weiting3

（1.School of Information Management，Wuhan University，Wuhan 430072，China;

2.School of Information Management，Central China Normal University，Wuhan 430079，China）

Abstract：[Purpose/Significance]Context-aware recommendation is an important stage in the development of the recommendation system.Its core is to integrate context information into the process of resource recommendation，in order to improve the accuracy of the resource recommendation system.As an important technic for resource recommendation，context-aware recommendation has attracted extensive attention in various fields and has gradually become an indispensable part of resource recommendation research.[Method/Process]This paper took the literature collections selected from WOS，CNKI and other databases as the research object.Firstly，it combed and summarized the main methods of context-aware recommendation systematically-pre-spatial filtering recommendation method，post-context filtering recommendation method and context modeling.Then this paper summarized and analyzed the progress of the context-aware recommendation in practice from two aspects：context data collection and context awareness application.Finally，it summarized the theoretical value and practical significance of context-awareness and context-awareness research on resource recommendation research.[Result/Conclusion]Future research would integrate three mainstream methods of the context-aware resource recommendation，integrate contextual elements in a mixed form and build a multidimensional and complex context-aware recommendation model，finally further enhance the information resource service experience in a refined and personalized way.

Key words：context-awareness;resource recommendation;context-aware recommendation

移動智能和物联网时代的开启，促使网络信息资源迅速增长，同时承载用户行为信息的传感器数据规模也在不断攀升。信息资源过载（Information Overload）现象愈发严峻，用户急需有效工具来快速、准确地获取所需信息资源。推荐系统通过分析和预测用户需求帮助用户实现信息过滤，因此逐渐成为解决信息过载问题的重要手段之一。自1994年明尼苏达大学研发基于用户的协同过滤资源推荐算法以来，推荐系统的研究已经历几十年的发展[1]。……