Conference Paper

Improving user interest inference from social neighbors.

DOI: 10.1145/2063576.2063720 Conference: Proceedings of the 20th ACM Conference on Information and Knowledge Management, CIKM 2011, Glasgow, United Kingdom, October 24-28, 2011
Source: DBLP

ABSTRACT Prior research has provided some evidence of social correlation (i.e., "you are who you know"), which makes it possible to infer one's interests from his or her social neighbors. However, it is also shown to be challenging to consistently obtain high quality inference. This challenge can be partially attributed to the fact that people usually maintain diverse social relationships, in order to tap into diverse information and knowledge. It is unlikely that a person would possess all interests of his/her social neighbors. Instead, s/he may selectively acquire just a subset of them. This paper intends to improve inferring interests from neighbors given this observation. We conduct this study by implementing a privacy-preserving large distributed social sensor system in a large global IT company to capture the multifaceted activities (e.g., emails, instant messaging, social bookmarking, etc.) of 25K+ people. These activities occupy the majority of employees' time, and thus, provide a higher quality view of the diverse aspects of their professional interests compared to the friending activity on online social networking sites. In this paper, we propose a technique that exploits the correlation among the attributes that a person possesses to improve social-correlation-based inference quality. Our technique offers two unique contributions. First, we demonstrate that the proposed technique can significantly improve inference quality by as much as 76.1%. Second, we study the interaction between the two factors: social correlation and attribute correlation under different situations. The results can inform practical applications how the inference quality would change in various scenarios.

  • [Show abstract] [Hide abstract]
    ABSTRACT: With the rapid development of social media applications, lots of users are connected with friends online, and their daily life and opinions are recorded. Social media provides us an unprecedented way to collect and analyze billions of users’ information. Proper user attribute identification or profile inference becomes increasingly attractive and feasible. However, the flourishing social records also pose great challenge in effective feature selection and integration for user profile inference, which is mainly caused by the text diversity and complex community structures. In this paper, we propose a comprehensive framework to infer the user occupation from his/her social activities recorded in the micro-blog system, which is a multi-source integration framework that combines both content and network information. We first identify some beneficial content features, and propose a machine learning classification model, named content model. We proceed to exploit the social network information, which tailors a community discovery based latent dimension solution to extract community-based feature, and utilizes the neighbor predictions for inference updating. Extensive empirical studies are conducted on a large real-life micro-blog dataset. The experimental results demonstrate the superiority of our integrated model for the occupation inference task, verify the effect of homophily in user interaction records, and reveal different effects of heterogeneous interactive networks.
    World Wide Web 01/2014; DOI:10.1007/s11280-014-0300-6 · 1.62 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Different users have different needs, it is increasingly difficult to recommend interested topics to them. The micro-blogging system can expose user interests from individual behaviors along with his/her social connections. It also offers an opportunity to investigate how a large-scale social system recommends personal preferences according to the temporal, spatial and topical aspects of users activity. Here we focus on the problem of mining user interest and modeling its evolution on the micro-blogging system for recommendation. We learn the user preference on topics from the visited micro-bloggings as user interest using text mining techniques. We then extend this concept with user's social connection on different topics. Moreover, we study the evolution of the user interest model and finally recommend the most preferred micro-bloggings to a user. Experiments on a large scale of micro-blogging dataset shows that our model outperforms traditional approaches and achieves considerable performance on recommending interested posts to a user.
    Proceedings of the 14th international conference on Web-Age Information Management; 06/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: Social recommender system has become an emerging research topic due to the prevalence of online social networking services during the past few years. In this paper, aiming at providing fundamental support to the research of social recommendation problem, we conduct an in-depth analysis on the correlations between social friend relations and user interest similarities. When evaluating interest similarities without distinguishing different friends a user has, we surprisingly observe that social friend relations generally cannot represent user interest similarities. A user's average similarity on all his/her friends is even correlated with the average similarity on some other randomly selected users. However, when measuring interest similarities using a finer granularity, we find that the similarities between a user and his/her friends are actually controlled by the network structure in the friend network. Factors that affect the interest similarities include subgraph topology, connected components, number of co-friends, etc. We believe our analysis provides substantial impact for social recommendation research and will benefit ongoing research in both recommender systems and other social applications.