ABSTRACT
In this research, we propose a Voter Model introducing expertise weights and friendshipness weights, which is an opinion formation model more suitable to the reality. Since the conventional Voter Model uniformly handles the following users, it is a model that adopts the opinion of the majority of follow users. When choosing an item from some alternatives, the adoption probability is determined by a majority vote. However, when actually making decisions, it is natural to consider the opinions of users who have expert knowledge about target products. Therefore, in this research, we propose a model that reflects the fact that the opinions of follow users are not handled equally, but the opinion of users with high specialty is preferentially selected. Specifically, we quantify the expertise of the user and introduce adoption probability proportional to expertise. Also, at the time of user's decision, it is natural to consider the opinion of users who have similar features such as profiles and circumstances with themselves. Therefore, in this research, we also propose a model reflecting the fact that users' opinions of users with high friendshipness are preferentially selected. In our experimental evaluation, actual data including user's following relationship, user profile, and categories of review items, at the cosmetic site @cosme are used. In the ranking of expected influence degree calculated by the proposed method, we quantitatively evaluate whether reviews of highly ranked users influence other users. We also evaluate how much the ranking result differs from the conventional Voter Model and centrality rankings.
- D. M. Blei, A. Y. Ng, and M. I. Jordan. 2003. Latent Dirichlet Allocation. J. Mach. Learn. Res. 3 (March 2003), 993--1022. Google ScholarDigital Library
- S. Brin and L. Page. 1998. The anatomy of a large-scale hypertextual web search engine. Computer Networks and ISDN Systems 30 (1998), 107--117. Google ScholarDigital Library
- E. Even-Dar and A. Shapira. 2007. A Note on Maximizing the Spread of Influence in Social Networks. In Internet and Network Economics, Xiaotie Deng and Fan Graham (Eds.). Lecture Notes in Computer Science, Vol. 4858. Springer Berlin / Heidelberg, 281--286. Google ScholarDigital Library
- T. Fushimi, K. Saito, M. Kimura, H. Motoda, and K. Ohara. 2010. Finding relation between PageRank and voter model. In Proceedings of the 11th international conference on Knowledge management and acquisition for smart systems and services (PKAW'10). Springer-Verlag, Berlin, Heidelberg, 208--222. Google ScholarDigital Library
- T. H. Haveliwala. 2003. Topic-Sensitive PageRank: A Context-Sensitive Ranking Algorithm for Web Search. IEEE Trans. on Knowl. and Data Eng. 15, 4 (July 2003), 784--796. Google ScholarDigital Library
- Masahiro Kimura, Kazumi Saito, Kouzou Ohara, and Hiroshi Motoda. 2011. Detecting Anti-majority Opinionists Using Value-weighted Mixture Voter Model. In Proceedings of the 14th International Conference on Discovery Science (DS'11). Springer-Verlag, Berlin, Heidelberg, 150--164. Google ScholarDigital Library
- Masahiro Kimura, Kazumi Saito, Kouzou Ohara, and Hiroshi Motoda. 2013. Learning to predict opinion share and detect anti-majority opinionists in social networks. Journal of Intelligent Information Systems 41 (08 2013), 5--37. Issue 1.Google Scholar
- A. N. Langville and C. D. Meyer. 2004. Deeper Inside PageRank. Internet Mathematics 1, 3 (2004), 335--380.Google ScholarCross Ref
- Y. Li, W. Chen, Y. Wang, and Z.-L. Zhang. 2015. Voter Model on Signed Social Networks. Internet Mathematics 11, 2 (2015), 93--133.Google ScholarCross Ref
- T. M. Liggett. 1999. Stochastic Interacting Systems: Contact, Voter and Exclusion Processes (Grundlehren der mathematischen Wissenschaften) (1 ed.). Springer.Google Scholar
- V. Sood, T. Antal, and S. Redner. 2008. Voter models on heterogeneous networks. Physical Review E 77, 4 (2008), 041121.Google ScholarCross Ref
- Yuki Yamagishi, Kazumi Saito, Kouzou Ohara, Masahiro Kimura, and Hiroshi Motoda. 2011. Learning Attribute-weighted Voter Model over Social Networks. In Proceedings of the 3rd Asian Conference on Machine Learning (ACML 2011). JMLR.org, 263--280.Google Scholar
Index Terms
- Modeling Opinion Formation by Incorporating Users' Expertise and Friendshipness
Recommendations
Detection of web users' opinion from multimodal opinion elements
COMPUTE '11: Proceedings of the Fourth Annual ACM Bangalore ConferenceIn this paper we present an approach to identify opinion of web users from opinionated texts and to classify web users opinion into positive or negative. Today web users express their opinion using multimodal opinion elements such as opinion phrases, ...
Role of conformity in opinion dynamics in social networks
COSN '14: Proceedings of the second ACM conference on Online social networksSocial networks serve as important platforms for users to express, exchange and form opinions on various topics. Several opinion dynamics models have been proposed to characterize how a user iteratively updates her expressed opinion based on her innate ...
Modelling majority and expert influences on opinion formation in online social networks
Two most important social influences that shape the opinion formation process are: (i) the majority influence caused by the existence of a large group of people sharing similar opinions and (ii) the expert influence originated from the presence of ...
Comments