Learning to blend vitality rankings from heterogeneous social networks

Abstract

Heterogeneous social network services, such as Facebook and Twitter, have emerged as popular, and often effective channels for Web users to capture updates from their friends. The explosion in popularity of these social network services, however, has created the problem of "information overload". The problem is becoming more severe as more and more users have engaged in more than one social networks simultaneously, each of which usually yields different friend connections and various sources of updates. Thus, it has made necessity to perform effective information filtering to retrieve information really attractive to web users from each of social networks and further blend them into a unified ranking list. In this paper, we introduce the problem of blending vitality rankings from heterogeneous social networks, where vitality denotes all kinds of updates user receives in various social networks. We propose a variety of content, users, and users correlation features for this task. Since vitalities from different social networks are likely to have different sets of features, we employ a divide-and-conquer strategy in order to fully exploit all available features for vitalities from each social network, respectively. Our experimental results, obtained from a large scale evaluation over two popular social networks, demonstrate the effectiveness of our method for putting vitalities that really interest users into higher orders in the blended ranking list. We complement our results with a thorough investigation of the feature importance and model selection with respect to both blending strategy and ranking for each social network.

Differences Between People Who Use Only Facebook and Those Who Use Facebook Plus Twitter

“…However, to our knowledge, there have been currently no studies exploring the differences in personality between those who use one of these two sites alone and those who use both FB and Twitter. This is relevant, considering that many users tend to be active in more than one SNS, a phenomena that has recently lead to the formation of the so called “social network aggregation platforms” (Bian, YiChang, YunFu, & Wen-YenChen, 2012). In order to examine this issue, we explored differences in FB and Twitter use, alone or in combination, in Narcissism, Loneliness, Shyness, and the Big Five Personality Traits (Emotional Stability, Agreeableness, Conscientiousness, Extraversion, and Intellect or Imagination) in a sample of college students.…”

“…This is relevant, considering that many users tend to be active in more than one SNS, a phenomena that has recently lead to the formation of the so called “social network aggregation platforms” (Bian, YiChang, YunFu, & Wen-YenChen, 2012). These aggregation services allow users to collect content from multiple social network services, pulling together information into a single location or consolidating multiple social networking profiles into one profile (Benevenuto, Rodrigues, Cha, & Almeida, 2012).…”

Aggregated Search

“…Bian et al [2012] proposed an algorithm for ranking social network updates originating from different networks. The main challenge is that different networks may be associated with different sources of evidence that can be used to predict the relevance of an update for a particular user.…”

“…Moreover, some features may only exist in one network and not the other (e.g., the number of Facebook chat messages between the user and the author of an update). Rather than rank candidate updates from different networks using a single model, Bian et al [2012] describe a Introduction "divide and conquer" approach that learns network-specific rankers and combines their output rankings into a single merged list. Lee et al [2012] focused on the task of ranking social media updates and used two test collections: one generated from Facebook updates and another generated from Twitter updates.…”

Dimensionality reduction of data sequences for human activity recognition

“…An important area of social network is exploring new methods for behavior analysis [4][5] [6]. However, explosion in popularity of social network services leads to the problem of "information overload" [7]. To let users more efficiently connect and communicate, it is necessary to introduce effective information extracting mechanism [8] [9][10] [11][12] to identify information most interesting to users from every network [13].…”