Learning to rank social update streams

Abstract

As online social media further integrates deeper into our lives, we spend more time consuming social update streams that come from our online connections. Although social update streams provide a tremendous opportunity for us to access information on-the-fly, we often complain about its relevance. Some of us are flooded with a steady stream of information and simply cannot process it in full. Ranking the incoming content becomes the only solution for the overwhelmed users. For some others, in contrast, the incoming information stream is pretty weak, and they have to actively search for relevant information which is quite tedious. For these users, augmenting their incoming content flow with relevant information from outside their first-degree network would be a viable solution. In that case, the problem of relevance becomes even more prominent.In this paper, we start an open discussion on how to build effective systems for ranking social updates from a unique perspective of LinkedIn -the largest professional network in the world. More specifically, we address this problem as an intersection of learning to rank, collaborative filtering, and clickthrough modeling, while leveraging ideas from information retrieval and recommender systems. We propose a novel probabilistic latent factor model with regressions on explicit features and compare it with a number of non-trivial baselines. In addition to demonstrating superior performance of our model, we shed some light on the nature of social updates on LinkedIn and how users interact with them, which might be applicable to social update streams in general.

Leveraging Social Connections to Improve Personalized Ranking for Collaborative Filtering

“…Many existing works aim to use social connections in order to predict the ratings that each user will give to each item; a commonly used assumption is that each user's rating behavior should be somehow similar to that of their friends. However, in many real applications, explicit numerical ratings might not be available and one must instead try to model some form of implicit feedback, such as the media they consume, the pages they browse, the music they listen to, or whom they befriend [4,21]. This setting is called "one-class" recommendation and a variety of solutions have been proposed to solve it by directly modeling relative preferences, or rankings, of items for personalized recommendation [8,12,18,22].…”

“…To motivate this work, we first conduct a simple analysis of preference data from four sources that we will consider in this paper: Ciao 1 , Delicious 2 , LibraryThing 3 and Epinions 4 . All four sources consist of preference data in addition to explicit social connections.…”

CoFiSet: Collaborative Filtering via Learning Pairwise Preferences over Item-sets

“…From Table 4 and discussions above, we can see that (1) CoFiSet is different from other algorithms, since it is based on a new assumption of pairwise preferences over item-sets, and (2) the most closely related works are BPRMF [17], CCF(SoftMax) [20] and PLMF [5], because they also adopt pairwise preference assumptions, exponential family functions in loss terms, and SGD (stochastic gradient descent) style algorithms.…”

“…However, in real applications, users' explicit ratings are not easily obtained, so they are not sufficient for the purpose of training an adequate prediction model, while users' implicit data like browsing and shopping records can be more easily collected. Some recent works have thus turned to improve the recommendation performance via exploiting users' implicit feedbacks, which include users' logs of clicking social updates [5], watching TV programs [6], assigning tags [14], purchasing products [17], browsing web pages [20], etc.…”

Ranking in information streams

“…Guy et al [4] propose user modelling techniques to improve a user's information stream. More recently, Hong et al [5] have looked at the Pairwise LTR technique to improve click-through rates on LinkedIn.…”