Link Prediction via Matrix Factorization

Menon, Aditya Krishna; Elkan, Charles

doi:10.1007/978-3-642-23783-6_28

Cited by 391 publications

(295 citation statements)

References 28 publications

Supporting

Mentioning

294

Contrasting

Order By: Relevance

“…One-class MF is a case of PU (positive-unlabeled) learning [10], which includes other important applications such as link prediction [17]. Our proposed optimization methods can be easily applied to them.…”

mentioning

confidence: 99%

Selection of Negative Samples for One-class Matrix Factorization

Yu¹,

Bilenko²

2017

Proceedings of the 2017 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

Many recommender systems have only implicit user feedback. The two possible ratings are positive and negative, but only part of positive entries are observed. One-class matrix factorization (MF) is a popular approach for such scenarios by treating some missing entries as negative. Two major ways to select negative entries are by sub-sampling a set with similar size to that of observed positive entries or by including all missing entries as negative. They are referred to as "subsampled" and "full" approaches in this work, respectively. Currently detailed comparisons between these two selection schemes on large-scale data are still lacking. One important reason is that the "full" approach leads to a hard optimization problem after treating all missing entries as negative. In this paper, we successfully develop e cient optimization techniques to solve this challenging problem so that the "full" approach becomes practically viable. We then compare in detail the two approaches "subsampled" and "full" for selecting negative entries. Results show that the "full" approach of including much more missing entries as negative yields better results.

show abstract

mentioning

confidence: 99%

Selection of Negative Samples for One-class Matrix Factorization

Yu¹,

Bilenko²

2017

Proceedings of the 2017 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

show abstract

“…For example, similar properties such as transitivity and homophily have been observed for positive links [99,77]; trust prediction algorithms perform well with positive links [57,77]; and many trust application frameworks can be directly applied to positive links [82,46]. Since distrust is a special type of negative links, a natural question here is whether we can generalize some properties and algorithms of distrust to negative links such as foes and dislike.…”

Section: Generalizing Findings Of Distrustmentioning

confidence: 89%

Computing Distrust in Social Media

Tang¹

2015

View full text Add to dashboard Cite

A myriad of social media services are emerging in recent years that allow people to communicate and express themselves conveniently and easily. The pervasive use of social media generates massive data at an unprecedented rate. It becomes increasingly difficult for online users to find relevant information or, in other words, exacerbates the information overload problem. Meanwhile, users in social media can be both passive content consumers and active content producers, causing the quality of user-generated content can vary dramatically from excellence to abuse or spam, which results in a problem of information credibility. Trust, providing evidence about with whom users can trust to share information and from whom users can accept information without additional verification, plays a crucial role in helping online users collect relevant and reliable information. It has been proven to be an effective way to mitigate information overload and credibility problems and has attracted increasing attention.As the conceptual counterpart of trust, distrust could be as important as trust and its value has been widely recognized by social sciences in the physical world. However, little attention is paid on distrust in social media. Social media differs from the physical world -(1) its data is passively observed, large-scale, incomplete, noisy and embedded with rich heterogeneous sources; and (2) distrust is generally unavailable in social media. These unique properties of social media present novel challenges for computing distrust in social media: (1) passively observed social media data does not provide necessary information social scientists use to understand distrust, how can I understand distrust in social media? (2) distrust is usually invisible in social media, how can I make invisible distrust visible by leveraging unique properties of social media data? and (3) little is known about distrust and its role in social media applications, how can distrust help make difference in social media applications?The chief objective of this dissertation is to figure out solutions to these challenges i via innovative research and novel methods. In particular, computational tasks are designed to understand distrust, a innovative task, i.e., predicting distrust is proposed with novel frameworks to make invisible distrust visible, and principled approaches are develop to apply distrust in social media applications. Since distrust is a special type of negative links, I demonstrate the generalization of properties and algorithms of distrust to negative links, i.e., generalizing findings of distrust, which greatly expands the boundaries of research of distrust and largely broadens its applications in social media. He is the mentor because I always seek his suggestions and help for many aspects in my life; and he is also my friend because I feel free to share my personal happiness and sadness with him. Dr.Liu, I cannot thank you enough.

show abstract

“…We used the area under the ROC curve (AUC) measure, which is not influenced by the imbalance distribution of the classes [10], for evaluation of different classification models. Additionally, we used the Precision and Recall measures in order to verify the ranking performance of our algorithm.…”

Section: Discussionmentioning

confidence: 99%

Intruder or Welcome Friend: Inferring Group Membership in Online Social Networks

Lesser

Tenenboim-Chekina

Rokach

et al. 2013

Social Computing, Behavioral-Cultural Modeling and Prediction

View full text Add to dashboard Cite

Abstract. Inferring Online Social Networks (OSN) group members may help to evaluate the authenticity of an applicant asking to join a certain group, and secure vulnerable populations online, such as children. We propose machine learning based methods, which associate OSN members' affiliation with virtual groups based on personal, topological, and group affiliation features. The study applies and evaluates the methods empirically, on two social networks (Ning and TheMarker). The experimental results demonstrate that one can accurately determine the group genuine members. Our study compares personal, topological and group based classification models. The results show that topological and group affiliation attributes contribute the most to group inference accuracy. Additionally, we examine the relations among the groups and identify group clustering tendencies where some groups are more tightly connected than others.

show abstract

Link Prediction via Matrix Factorization

Cited by 391 publications

References 28 publications

Selection of Negative Samples for One-class Matrix Factorization

Selection of Negative Samples for One-class Matrix Factorization

Computing Distrust in Social Media

Intruder or Welcome Friend: Inferring Group Membership in Online Social Networks

Contact Info

Product

Resources

About