基于语义行为和社交关联的好友推荐模型

doi:10.13232/j.cnki.jnju.2018.06.008

南京大学学报(自然科学版) ›› 2018, Vol. 54 ›› Issue (6): 1124–1131.doi: 10.13232/j.cnki.jnju.2018.06.008

基于语义行为和社交关联的好友推荐模型

曲昭伟^１，杨凯翔³*，王晓茹²，李百威^１

1. 北京邮电大学网络技术研究院，北京，100876；2. 北京邮电大学计算机学院，北京，100876； 3. 吉林建筑大学土木工程学院，长春，130118

接受日期:2018-10-15 出版日期:2018-12-01 发布日期:2018-12-01
通讯作者: 杨凯翔, yvf5242172jitan@163.com E-mail:yvf5242172jitan@163.com
基金资助:
国家自然科学基金(61672108)

Friend recommendation model based on semantic behavior and social links

Qu Zhaowei¹，Yang Kaixiang³*，Wang Xiaoru²，Li Baiwei¹

1.Institute of Network Technology，Beijing University of Posts and Telecommunication，Beijing，100876，China； 2．School of Computer Science，Beijing University of Posts and Telecommunication，Beijing，100876，China； 3．College of Civil Engineering，Jilin University of Architecture，Changchun，130118，China

Accepted:2018-10-15 Online:2018-12-01 Published:2018-12-01
Contact: Yang Kaixiang, yvf5242172jitan@163.com E-mail:yvf5242172jitan@163.com

摘要/Abstract

摘要： 现今社交媒体是建立社交联系的重要媒介，好友推荐对于扩展人们的关系网络起到至关重要的作用，准确的用户特征提取和分析是社交网络中好友推荐的关键. 传统的好友推荐方法一般都是根据部分用户属性信息或行为信息进行推荐，所以对用户特征的描述不完整，推荐的效率和准确率远非预期. 提出基于用户语义行为和社交关联的推荐模型应用于社交媒体平台上的好友推荐. 为了获得准确的预测，使用LDA(Latent Dirichlet Allocation)对语义信息进行主题建模，得到基于主题的用户语义行为特征表达；使用DeepWalk算法对用户社交关联网络图进行特征提取，得到准确的社交关联特征表达；使用反向传播神经网络来预测用户潜在的社交关联，为用户精准推荐好友. 该模型实现了利用用户语义行为和社交关联预测用户潜在的社交关联，可以根据潜在社交关联进行精准的好友推荐．

关键词: 语义行为, 社交关联, 主题分布, 好友推荐, 反向传播神经网络

Abstract: Nowadays，people usually establish social links on social media platforms，in which case friend recommendation plays a crucial role in expanding their communication. Accurate user feature extraction and analysis is the key to friend recommendation in social networks. Traditional methods of friend recommendation usually use some of users’ attribute information or behavior information which leads to the incomplete feature extraction，and the efficiency and accuracy of the recommendation are far from the expectation. We propose a recommendation model based on users’ semantic behavior and social links applied to the friend recommendation on social media platforms. In order to obtain accurate predictions，LDA(Latent Dirichlet Allocation) is used to deal with the semantic information to get topic－based representation of semantic behavior. DeepWalk algorithm is used to extract features of the graph which consists of social links to obtain accurate feature representations. The back propagation neural network is processed to predict the potential social links among users and recommend friends accurately. The model predicts the potential social links considering the semantic behavior and social links，and then recommends friends accurately with the potential social links.

Key words: semantic behavior, social links, topic distribution, friend recommendation, back－propagation neural network

中图分类号:

TP181

曲昭伟, 杨凯翔, 王晓茹, 李百威. 基于语义行为和社交关联的好友推荐模型[J]. 南京大学学报(自然科学版), 2018, 54(6): 1124–1131.

Qu Zhaowei, Yang Kaixiang, Wang Xiaoru, Li Baiwei. Friend recommendation model based on semantic behavior and social links[J]. Journal of Nanjing University(Natural Sciences), 2018, 54(6): 1124–1131.

参考文献

[1]　Ren C，An N，Wang J Z，et al. Optimal parameters selection for BP neural network based on particle swarm optimization：A case study of wind speed forecasting. Knowledge－Based Systems，2014，56：226－239.
[2] Errico J H，Sezan M I，Borden G R，et al. Collaborative recommendation system：U.S. Patent 8949899. 2015－02－03.
[3] Chin A. Finding cohesive subgroups and relevant members in the nokia friend view mobile social network ∥ International Conference on Computational Science and Engineering. Vancouver，Canada：IEEE，2009：278－283.
[4] Yang D Q，Zhang D Q，Yu Z Y，et al. A sentiment－enhanced personalized location recommendation system ∥ Proceedings of the 24th ACM Conference on Hypertext and Social Media. New York，NY，USA：ACM，2013：119－128.
[5] Lo S C，Lin C C. WMR－A graph－based algorithm for friend recommendation ∥ 2006 IEEE/WIC/ACM International Conference on Web Intelligence. Hong Kong，China：IEEE，2006：121－128.
[6] Huang S R，Zhang J，Wang L，et al. Social friend recommendation based on multiple network correlation. IEEE Transactions on Multimedia，2016，18(2)：287－299.
[7] He J N，Liu H Y，Lau R Y K，et al. Relationship identification across heterogeneous online social networks. Computational Intelligence，2017，33(3)：448－477.
[8] Cui L Z，Huang W Y，Yan Q，et al. A novel context－aware recommendation algorithm with two－level SVD in social networks. Future Generation Computer Systems，2017，86：1459－1470.
[9] Thijs G，Marchal K，Lescot M，et al. A Gibbs sampling method to detect over－represented motifs in the upstream regions of co－expressed genes ∥ Proceedings of the 5th Annual International Conference on Computational Biology. New York，NY，USA：ACM，2001：305－312.
[10] Perozzi B，Al－Rfou R，Skiena S. DeepWalk：online learning of social representations ∥ Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York，NY，USA：ACM，2014：701－710.
[11] Rong X. word2vec parameter learning explained. arXiv：1411.2738，2014.
[12] Lazaridou A，Pham N T，Baroni M. Combining language and vision with a multimodal skip－gram model. arXiv：1501.02598，2015.
[13] Fu K，Chan C H，Chau M. Assessing censorship on microblogs in China：Discriminatory keyword analysis and impact evaluation of the real name registration policy. IEEE Internet Computing，2013，17(3)：42－50.
[14] Zhang H P，Yu H K，Xiong D Y，et al. HHMM－based Chinese lexical analyzer ICTCLAS ∥ Sighan Workshop on Chinese Language Processing. Hong Kong，China：Morgan Kaufmann，2003：758－759.
[15] Lumer E. Social graph based recommender：WIPO Patent Application WO/2010/048172，2009－10－20.
[16] Liu J T，Wu C H，Wang J W. Gated recurrent units based neural network for time heterogeneous feedback recommendation. Information Sciences，2018，423：50－65.
[17] Chai T，Draxler R R. Root mean square error(RMSE)or mean absolute error(MAE)？－Arguments against avoiding RMSE in the literature. Geoscientific Model Development，2014，7(3)：1247－1250.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于语义行为和社交关联的好友推荐模型

Friend recommendation model based on semantic behavior and social links

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 8

Metrics

本文评价

推荐阅读 2

[1]	汪敏,赵飞,闵帆. 储层预测的代价敏感主动学习算法[J]. 南京大学学报(自然科学版), 2020, 56(4): 561-569.
[2]	朱荀,刘国强,丁华平,沈庆宏. 一种通过支持向量机对交通拥堵情况进行分类的方法[J]. 南京大学学报(自然科学版), 2020, 56(2): 278-283.
[3]	王卫星,刘兆伟,石敬华. 基于时间敏感滑动窗口的CP⁃nets结构学习[J]. 南京大学学报(自然科学版), 2020, 56(2): 175-185.
[4]	信统昌,刘兆伟. 基于贝叶斯⁃遗传算法的多值无环CP⁃nets学习[J]. 南京大学学报(自然科学版), 2020, 56(1): 74-84.
[5]	郑文萍,刘韶倩,穆俊芳. 一种基于相对熵的随机游走相似性度量模型[J]. 南京大学学报(自然科学版), 2019, 55(6): 984-999.
[6]	黄华娟,韦修喜. 基于自适应调节极大熵的孪生支持向量回归机[J]. 南京大学学报(自然科学版), 2019, 55(6): 1030-1039.
[7]	刘　素, 刘惊雷. 基于特征选择的CP－nets结构学习[J]. 南京大学学报(自然科学版), 2019, 55(1): 14-28.
[8]	贾海宁, 王士同. 面向重尾噪声的模糊规则模型[J]. 南京大学学报(自然科学版), 2019, 55(1): 61-72.