摘要
【目的/意义】帮助科研工作者了解领域相关科研工作者的研究内容,促进开展进一步的交流合作。【方法/过程】根据稀疏分布式表征理论对作者论文的文本内容进行特征表示,计算作者研究内容相似性,根据得分进行排序推荐。【结果/结论】选取了NIPS会议论文数据集进行了实验,将实验结果与TD-IDF方法和LDA主题模型方法进行对比,证明了稀疏分布式表征是一种能够从语义层面表征对象特征的数据表示方式,能有效用于合作推荐。
【Purpose/significance】Help scientific researchers understand the research content of relevant field scientists and promote further exchanges and collaboration.【Method/process】According to the sparse distributed representation theory, the text contents of the authors' thesis is characterized,and calculate the similarity of the author's research contents,sort by the score, then recommend.【Results/conclusion】The NIPS conference paper data set is selected for experiments. The experimental results are compared with the TD-IDF method and the LDA method, which proved that sparse distributed representation is a data representation method that can characterize the characteristics of objects from the semantic level and can be effectively used for collaborative recommendation.
引文
1 刘璇,朱庆华,段宇锋.社会网络分析法运用于科研团队发现和评价的实证研究[J].信息资源管理学报, 2011,1(3):32-37.
2 杨辰.科研社交网络平台中的合作者推荐[D].合肥:中国科学技术大学, 2015.
3 Hara N, Solomon P, Kim S L. An emerging view of scientific collaboration:Scientists’ perspectives on collaboration and factors that impact collaboration[J].Journal of the Association for Information Science&Technology, 2003, 54(10):952-965.
4 Kanfer A, Sweet J, Schlosser A. Humanizing the Net:Social Navigation with a“know-who”Email Agent[C]//Proceedings of the Third Conference on Human Factors and the Web. New York:ACM,1997.
5 张伟,黄炜,夏利民.基于广义内容概率潜在语义分析模型的推荐[J].计算机应用,2013,33(5):1330-1333.
6 Blei D M, Ng A Y, Jordan M I. Latent dirichlet allocation[J]. J Machine Learning Research Archive, 2003,(3):993-1022.
7 陈永恒,左万利,林耀进.作者标签主题模型在科技文献中的应用[J].计算机应用, 2015,35(4):1001-1005.
8 Webber F D S. Semantic Folding Theory And its Application in Semantic Fingerprinting[R]. Austria:Cortical.io, 2015.
9 吕琳媛.链路预测[M].北京:高等教育出版社, 2013.
10 Lee D H, Brusilovsky P, Schleyer T. Recommending collaborators using social features and MeSH terms[J]. Proceedings of the American Society for Information Science&Technology, 2011, 48(1):1-10.
11 邸亮,杜永萍. LDA模型在微博用户推荐中的应用[J].计算机工程, 2014, 40(5):1-6.
12 张亮.基于LDA主题模型的标签推荐方法研究[J].现代情报, 2016, 36(2):53-56.
13 刘萍,郑凯伦,邹德安.基于LDA模型的科研合作推荐研究[J].情报理论与实践, 2015, 38(9):79-85.
14 Fang H, Zhai C X. Probabilistic Models for Expert Finding[M]. Berlin:Springer Berlin Heidelberg, 2007.
15 Balog K, Azzopardi L, Rijke M D. Formal models for expert finding in enterprise corpora[C]//Proceedings of the 29 th annual international ACM SIGIR conference on Research and development in information retrieval, Seattle,Washington, USA:ACM New York, 2006:43–50.
16 Deng H, King I, Lyu M R. Formal Models for Expert Finding on DBLP Bibliography Data[C]//Eighth IEEE International Conference on Data Mining. New York:IEEE,2008:163-172.
17 陈卫静.基于合著网络的潜在合作关系挖掘[D].北京:中国科学院大学, 2013.
18 汪俊,岳峰,王刚等.科研社交网络中基于链接预测的专家推荐研究[J].情报杂志, 2015,(6):151-157.
19 熊回香,杨雪萍,蒋武轩,等.科研社交网站中基于相似兴趣的学者推荐研究[J].情报科学, 2017,V35(9):3-11.
20 霍金斯.人工智能的未来[M].西安:陕西科学技术出版社, 2006.
21 Ahmad S, Hawkins J. Properties of Sparse Distributed Representations and their Application to Hierarchical Temporal Memory[J]. Eprint Arxiv, 2015,(2):21-30.
22 付媛,朱礼军,韩红旗.姓名消歧方法研究进展[J].情报工程,2016,2(1):53-58.
23 Yuan Q, Zhao S, Chen L et al. Augmenting Collaborative Recommender by Fusing Explicit Social Relationships[C].Conference on Recommender Systems. New Yorl:ACM,2009.
24 Wind D K, M?rup M. Link prediction in weighted networks[C]//IEEE International Workshop on Machine Learning for Signal Processing. New York:IEEE, 2012:1-6.