基于三层维度的文献个性化推荐模型研究
|
摘要
【目的/意义】大数据情报分析和知识服务时代,如何快速高效地从海量文献中获取情报并实现精准的文献个性化推荐,是文献推荐个性化服务亟待解决的问题。【方法/过程】对文献个性化推荐模型进行研究,通过专家权重维、用户维以及情境感知维三个维度的协同,识别用户的兴趣点。推荐模型使用层次分析法和熵权法量化专家意见;使用潜在狄利克雷分布和KL散度计算量化用户相似度;通过用户社会标注行为、搜索行为、浏览行为得到用户情感倾向,并引入时间因子量化用户情感;最后引入"最大频度值"确定各个维度的推荐指数,加权计算得到文献综合推荐指数。【结果/结论】以高校图书馆为实验平台,对本文提出文献个性化推荐方法进行验证。实验结果表明,与传统的基于内容的推荐方法、协同过滤推荐方法以及混合的推荐方法相比,基于三层维度的文献个性化推荐方法在准确率与召回率上都取得了更好的性能。
【Purpose/significance】 In the era of big data intelligence analysis and knowledge service, it is an emergency issue to increase the effectiveness of intelligence obtaining and to achieve personalized precise literature recommendation service.【Method/process】Building a model from the perspective of three-layer dimension, which is the expert opinion dimension, user dimension and emotional perception dimension, to identify the interests of users. Using AHP and Entropy methodto calculate the expert opinion,using LDA and KL divergence to calculate the similarity of the users. Users' social senti-ment behaviors, search behaviors and browsing behaviors are used to derive user sentiment trends, the time factors are added to calculate user sentiment. Finally, by introducing"maximum frequency value", the index of the comprehensive literature recommendation are obtained to determine the recommend indexes.【Result/conclusion】Taking the university libraryas an example, the experiment to identify the result of our method is conducted. Compared with the traditional con-tent-based recommendation method, collaborative filtering recommendation method and mixed recommendation method,the proposed method improves the precision and recall index, and achieves better performance of personalized literature recommendations services.
【Purpose/significance】 In the era of big data intelligence analysis and knowledge service, it is an emergency issue to increase the effectiveness of intelligence obtaining and to achieve personalized precise literature recommendation service.【Method/process】Building a model from the perspective of three-layer dimension, which is the expert opinion dimension, user dimension and emotional perception dimension, to identify the interests of users. Using AHP and Entropy methodto calculate the expert opinion,using LDA and KL divergence to calculate the similarity of the users. Users' social senti-ment behaviors, search behaviors and browsing behaviors are used to derive user sentiment trends, the time factors are added to calculate user sentiment. Finally, by introducing"maximum frequency value", the index of the comprehensive literature recommendation are obtained to determine the recommend indexes.【Result/conclusion】Taking the university libraryas an example, the experiment to identify the result of our method is conducted. Compared with the traditional con-tent-based recommendation method, collaborative filtering recommendation method and mixed recommendation method,the proposed method improves the precision and recall index, and achieves better performance of personalized literature recommendations services.
引文
1熊回香,李跃艳.基于加权派系的个性化信息推荐研究[J].情报科学,2018,36(1):67-74.
2 Wang M X, Wang J Q, Li L. New online personalized recommendation approach based on the perceived value of consumer characteristics[J].Journal of Intelligent&Fuzzy Systems, 2017, 33(2):1-16.
3 Campigotto P, Rudloff C, Leodolter M, et al. Personalized and Situation-Aware Multimodal Route Recommendations:The FAVOUR Algorithm[J].IEEE Transactions on Intelligent Transportation Systems,2017, 18(1):92-102.
4 Kuai H, Yan J, Chen J, et al. A knowledge-driven approach for personalized literature recommendation based on deep semantic discrimination[C]//International Conference on Web Intelligence. ACM, 2017:1253-1259.
5 Wang N, Zhong N, Han J, et al. A Personalized Method of Literature Recommendation Based on Brain Informatics Provenances[C]//International Conference on Brain Informatics and Health. Springer,Cham, 2015:167-178.
6 Guan P, Wang Y. Personalized scientific literature re commendation based on user's research interestC]//International Conference on Natural Computation,Fuzzy Systems and Knowledge Discovery. IEEE,2016:1273-1277.
7 Aleksandra Kla?nja-Milicevic, et al. Enhancing e-lear ning systems with personalized recommendation based on collaborative tagging techniques[J].Applied Intelligence, 2017,(1):1-17.
8 Wang C, Zheng Y, Jiang J, et al. Toward Privacy-Preserving Personalized Recommendation Services[J].Engineering, 2018, 4(1):21-28.
9 MaK,LuT,AbrahamA.Hybrid parallel approach for personalized literature recommendation systemC]//International Conference on Computational Aspects of Social Networks. IEEE,2014:31-36.
10 Tewari A S, Barman A G. Sequencing of items in personalized recommendations using multiple recommendation techniques[J].Expert Systems with Applications, 2018,(97):70-82.
11 张麒麟,姜霖.基于文献内容的图书推荐机制研究[J].图书馆学研究,2018,(1):78-81,17.
12 王欣,张冬梅.大数据环境下基于高校用户小数据的图书馆个性化智能服务研究[J].情报理论与实践,2018,41(2):132-137.
13 Fattahi R, Khalilzadeh M. Risk evaluation using a novel hybrid method based on FMEA, extended MULTIMOORA,and AHP methods under fuzzy environment[J].Safety Science, 2018,(102):290-300.
14 Saaty T L. Analytic hierarchy process[J].Management Science and Engineering,2001,(9):109-121.
15 曲佳彬,欧石燕.基于主题过滤与主题关联的学科主题演化分析[J].数据分析与知识发现,2018,2(1):64-75.
16 王丽,邹丽雪,刘细文.基于LDA主题模型的文献关联分析及可视化研究[J].数据分析与知识发现,2018,2(3):98-106.
17 唐晓波,童海燕,严承希.基于话题情感强度的微博舆情分析[J].图书馆学研究,2014,(17):85-93.
18 侯银秀,李伟卿,王伟军,张婷婷.基于用户偏好与商品属性情感匹配的图书个性化推荐研究[J].数据分析与知识发现,2017,1(8):9-17.
19 朱娟,唐晓波.基于三层知识融合模型的个性化商品推荐[J].图书馆学研究,2017,(5):24-30.
20 王敏,嵇绍春.基于模糊聚类和模糊模式识别的数字图书馆个性化推荐研究[J].现代情报,2016,36(4):52-56.
21 郐弘智,闫健卓,陈建辉.一种知识驱动的个性化文献推荐方法[EB/OL].http://kns.cnki.net/kcms/detail/51.1196.TP.20171212.1834.074.html,2018-05-13.
2 Wang M X, Wang J Q, Li L. New online personalized recommendation approach based on the perceived value of consumer characteristics[J].Journal of Intelligent&Fuzzy Systems, 2017, 33(2):1-16.
3 Campigotto P, Rudloff C, Leodolter M, et al. Personalized and Situation-Aware Multimodal Route Recommendations:The FAVOUR Algorithm[J].IEEE Transactions on Intelligent Transportation Systems,2017, 18(1):92-102.
4 Kuai H, Yan J, Chen J, et al. A knowledge-driven approach for personalized literature recommendation based on deep semantic discrimination[C]//International Conference on Web Intelligence. ACM, 2017:1253-1259.
5 Wang N, Zhong N, Han J, et al. A Personalized Method of Literature Recommendation Based on Brain Informatics Provenances[C]//International Conference on Brain Informatics and Health. Springer,Cham, 2015:167-178.
6 Guan P, Wang Y. Personalized scientific literature re commendation based on user's research interestC]//International Conference on Natural Computation,Fuzzy Systems and Knowledge Discovery. IEEE,2016:1273-1277.
7 Aleksandra Kla?nja-Milicevic, et al. Enhancing e-lear ning systems with personalized recommendation based on collaborative tagging techniques[J].Applied Intelligence, 2017,(1):1-17.
8 Wang C, Zheng Y, Jiang J, et al. Toward Privacy-Preserving Personalized Recommendation Services[J].Engineering, 2018, 4(1):21-28.
9 MaK,LuT,AbrahamA.Hybrid parallel approach for personalized literature recommendation systemC]//International Conference on Computational Aspects of Social Networks. IEEE,2014:31-36.
10 Tewari A S, Barman A G. Sequencing of items in personalized recommendations using multiple recommendation techniques[J].Expert Systems with Applications, 2018,(97):70-82.
11 张麒麟,姜霖.基于文献内容的图书推荐机制研究[J].图书馆学研究,2018,(1):78-81,17.
12 王欣,张冬梅.大数据环境下基于高校用户小数据的图书馆个性化智能服务研究[J].情报理论与实践,2018,41(2):132-137.
13 Fattahi R, Khalilzadeh M. Risk evaluation using a novel hybrid method based on FMEA, extended MULTIMOORA,and AHP methods under fuzzy environment[J].Safety Science, 2018,(102):290-300.
14 Saaty T L. Analytic hierarchy process[J].Management Science and Engineering,2001,(9):109-121.
15 曲佳彬,欧石燕.基于主题过滤与主题关联的学科主题演化分析[J].数据分析与知识发现,2018,2(1):64-75.
16 王丽,邹丽雪,刘细文.基于LDA主题模型的文献关联分析及可视化研究[J].数据分析与知识发现,2018,2(3):98-106.
17 唐晓波,童海燕,严承希.基于话题情感强度的微博舆情分析[J].图书馆学研究,2014,(17):85-93.
18 侯银秀,李伟卿,王伟军,张婷婷.基于用户偏好与商品属性情感匹配的图书个性化推荐研究[J].数据分析与知识发现,2017,1(8):9-17.
19 朱娟,唐晓波.基于三层知识融合模型的个性化商品推荐[J].图书馆学研究,2017,(5):24-30.
20 王敏,嵇绍春.基于模糊聚类和模糊模式识别的数字图书馆个性化推荐研究[J].现代情报,2016,36(4):52-56.
21 郐弘智,闫健卓,陈建辉.一种知识驱动的个性化文献推荐方法[EB/OL].http://kns.cnki.net/kcms/detail/51.1196.TP.20171212.1834.074.html,2018-05-13.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |