针对现有微博推荐方法或模型不便组合证据的不足,提出一种微博推荐新方法.将信念网络用于微博推荐,构建一个基本信念网络推荐模型,并在基本模型中融合用户交互微博证据,提出一个微博推荐扩展模型.模型中的节点表示微博、用户和兴趣特征词,有向弧表示节点之间的关系,通过计算用户与微博的覆盖程度来得到用户与待评估微博的相关度.在微博数据集上的实验结果表明:扩展模型较基本模型在F值上至少提高了约4.9%;与已有的推荐方法相比,新模型在组合证据提高推荐性能方面更有效.
Considering that the existing microblog recommendation methods or models are inconvenient to combine evidence, this paper proposes a new microblog recommendation method. The belief network is applied to microblog recommendation, a basic belief network recommendation model is then constructed, and an extended microblog recommendation model is proposed by integrating the user interaction microblog evidence into the basic model. The nodes in the model represent microblogs, users and interest feature words, while the directed arcs represent relationships between nodes. The relevance between user and microblog is obtained by calculating the degree of coverage between user and microblog. The experimental results on microblog datasets show that the F-measure of the extended model is at least 4.9% higher than that of the basic model. Compared with existing recommendation methods, the new model is more effective in improving recommendation performance in terms of combining evidence.
[1] 于亚新,刘梦,张宏宇.Twitter社交网络用户行为理解及个性化服务推荐算法研究[J].计算机研究与发展, 2020, 57(7): 1369-1380. DOI: 10.7544/issn1000-1239.2020.20190158.
[2] 新浪科技.微博2021年第三季度财报[EB/OL]. [2021-11-12]. https://weibo.com/1642634100//L10O1BwkR.2021.11.
[3] GREWAL A, LIN J. The evolution of content analysis for personalized recommendations at Twitter[C] //Proceedings of the 41th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2018: 1355-1356. DOI: 10.1145/3209978.3210206.
[4] CHEN L, LYU D D, XU Z X, et al. A content-location-aware public welfare activity information push system based on microblog[J]. Information Processing and Management, 2020, 57(1): 102137.1-102137.12. DOI: 10.1016/j.ipm.2019.102137.
[5] WANG Y R, DING S F, XU X, et al. The multi-tag semantic correlation used for micro-blog user interest modeling[J]. Engineering Applications of Artificial Intelligence, 2019, 85(10): 765-772. DOI: 10.1016/j.engappai.2019.08.007.
[6] 马慧芳,张迪,赵卫中,等.基于超图随机游走标签扩充的微博推荐方法[J].软件学报, 2019, 30(11): 3397-3412. DOI: 10.13328/j.cnki.jos.005545.
[7] SUN A X. Short text classification using very few words[C] //Proceedings of the 35th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2012: 1145-1146. DOI: 10.1145/2348283.2348511.
[8] 王战平,夏榕.基于社会化标签挖掘的微博内容推荐方法研究[J].情报科学, 2021, 39(5): 91-96. DOI: 10.13833/j.issn.1007-7634.2021.05.013.
[9] 王臻皇,陈思明,袁晓如.面向微博主题的可视分析研究[J].软件学报, 2018, 29(4): 1115-1130. DOI: 10.13328/j.cnki.jos.005261.
[10] 邢千里,刘列,刘奕群,等.微博中用户标签的研究[J].软件学报, 2015, 26(7): 1626-1637. DOI: 10.13328/j.cnki.jos.004655.
[11] 郑伟,侯宏旭,班志杰.贝叶斯网络查询语义扩展的专家发现方法[J].计算机工程与应用, 2020, 56(13): 194-198. DOI: 10.3778/j.issn.1002-8331.1903-0184.
[12] LI C Y, ZHANG Y S. A personalized recommendation algorithm based on large-scale real micro-blog data[J]. Neural Computing and Applications, 2020, 32(1): 11245-11252. DOI: 10.1007/s00521-020-05042-y.
[13] ZHANG S X, LIU W J, DENG X L, et al. Micro-blog topic recommendation based on knowledge flow and user selection[J]. Journal of Computational Science, 2018, 26(5): 512-521. DOI: 10.1016/j.jocs.2017.10.021.
[14] KIM Y, SHIM K. Twitobi: A recommendation system for twitter using probabilistic modeling[C] //The 11th IEEE International Conference on Data Mining, IEEE, 2011: 340-349. DOI: 10.1109/ICDM.2011.150.
[15] 刘宇东,孙豪,蒋运承.融合内容相似度与多特征计算的个性化微博推荐模型[J].计算机科学, 2020, 47(10): 97-101. DOI: 10.11896/jsjkx.190700073.
[16] 陈杰,刘学军,李斌,等.一种基于用户动态兴趣和社交网络的微博推荐方法[J]. 电子学报, 2017, 45(4): 898-905. DOI: 10.3969/j.issn.0372-2112.2017.04.019.
[17] 韩康康,徐建民,张彬.融合用户兴趣和多维信任度的微博推荐[J].数据分析与知识发现, 2020, 4(12): 95-104. DOI: 10.11925/infotech.2096-3467.2020.0049.
[18] 赵森栋,刘挺.因果关系及其在社会媒体上的应用研究综述[J].软件学报, 2014, 25(12): 2733-2752. DOI: 10.13328/j.cnki.jos.004724.
[19] 徐建民.基于术语关系的贝叶斯网络检索模型扩展[M].北京: 科学出版社, 2019: 80-83.
[20] PAN H L, YANG X H. Intelligent recommendation method integrating knowledge graph and Bayesian network[J]. Soft Computing, 2021: 1-10. DOI: 10.1007/s00500-021-05735-z.
[21] HUANG J J, ZHU K L, ZHONG N. A probabilistic inference model for recommender systems[J]. Applied Intelligence, 2016, 45(3): 686-694. DOI: 10.1007/s10489-016-0783-1.
[22] CALADO P, RIBEIRO-NETO B, ZIVIANI N, et al. Local versus global link information in the Web[J]. ACM Transactions on Information Systems, 2003, 21(1): 42-63. DOI: 10.1145/635484.635486.
[23] 石磊,陶永才,李俊艳,等.个性化微博实时推荐模型研究[J].小型微型计算机系统, 2016, 37(9): 1910-1914. DOI: CNKI:SUN:XXWX.0.2016-09-004.
[24] 印鉴,王智圣,李琪,等.基于大规模隐式反馈的个性化推荐[J].软件学报, 2014, 25(9): 1953-1966. DOI: 10.13328/j.cnki.jos.004648.
[25] CRISTO M A P D, CALADO P P, SILVEIRA M D L D, et al. Bayesian belief networks for IR[J]. International Journal of Approximate Reasoning, 2003, 34(2-3): 163-179. DOI: 10.1016/j.ijar.2003.07.006.
[26] 朱郁筱,吕琳媛.推荐系统评价指标综述[J].电子科技大学学报, 2012, 41(2): 163-175. DOI: 10.3969/j.issn.1001-0548.2012.02.001.
[27] LANDIS J R, KOCH G G. The measurement of observer agreement for categorical data[J]. Biometrics, 1977, 33(3): 159-174. DOI: 10.2307/2529310. (
