Loading [MathJax]/extensions/tex2jax.js

Real-time top-N recommendation in social streams

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Authors

Research Organisations

External Research Organisations

  • University of Hildesheim

Details

Original languageEnglish
Title of host publicationRecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems
Pages59-66
Number of pages8
Publication statusPublished - 9 Sept 2012
Event6th ACM Conference on Recommender Systems, RecSys 2012 - Dublin, Ireland
Duration: 9 Sept 201213 Sept 2012

Publication series

NameRecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems

Abstract

The SocialWeb is successfully established, and steadily growing in terms of users, content and services. People generate and consume data in real-time within social networking services, such as Twitter, and increasingly rely upon continuous streams of messages for real-time access to fresh knowledge about current affairs. In this paper, we focus on analyzing social streams in real-time for personalized topic recommen- dation and discovery. We consider collaborative filtering as an online ranking problem and present Stream Ranking Matrix Factorization { RMFX {, which uses a pairwise approach to matrix factorization in order to optimize the personalized ranking of topics. Our novel approach follows a selective sampling strategy to perform online model updates based on active learning principles, that closely simulates the task of identifying relevant items from a pool of mostly uninteresting ones. RMFX is particularly suitable for large scale applications and experiments on the 476 million Twitter tweets dataset show that our online approach largely outperforms recommendations based on Twitter's global trend, and it is also able to deliver highly competitive Top-N recommendations faster while using less space than Weighted Regularized Matrix Factorization (WRMF), a state-of-the-art matrix factorization technique for Collaborative Filtering, demonstrating the efficacy of our approach.

Keywords

    Collaborative filtering, Matrix factorization, Online learning, Ranking, Selective sampling, Twitter

ASJC Scopus subject areas

Cite this

Real-time top-N recommendation in social streams. / Diaz-Aviles, Ernesto; Drumond, Lucas; Schmidt-Thieme, Lars et al.
RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems. 2012. p. 59-66 (RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems).

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Diaz-Aviles, E, Drumond, L, Schmidt-Thieme, L & Nejdl, W 2012, Real-time top-N recommendation in social streams. in RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems. RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems, pp. 59-66, 6th ACM Conference on Recommender Systems, RecSys 2012, Dublin, Ireland, 9 Sept 2012. https://doi.org/10.1145/2365952.2365968
Diaz-Aviles, E., Drumond, L., Schmidt-Thieme, L., & Nejdl, W. (2012). Real-time top-N recommendation in social streams. In RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems (pp. 59-66). (RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems). https://doi.org/10.1145/2365952.2365968
Diaz-Aviles E, Drumond L, Schmidt-Thieme L, Nejdl W. Real-time top-N recommendation in social streams. In RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems. 2012. p. 59-66. (RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems). doi: 10.1145/2365952.2365968
Diaz-Aviles, Ernesto ; Drumond, Lucas ; Schmidt-Thieme, Lars et al. / Real-time top-N recommendation in social streams. RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems. 2012. pp. 59-66 (RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems).
Download
@inproceedings{6679be6f669e4717ad1de44b1c7fa5cd,
title = "Real-time top-N recommendation in social streams",
abstract = "The SocialWeb is successfully established, and steadily growing in terms of users, content and services. People generate and consume data in real-time within social networking services, such as Twitter, and increasingly rely upon continuous streams of messages for real-time access to fresh knowledge about current affairs. In this paper, we focus on analyzing social streams in real-time for personalized topic recommen- dation and discovery. We consider collaborative filtering as an online ranking problem and present Stream Ranking Matrix Factorization { RMFX {, which uses a pairwise approach to matrix factorization in order to optimize the personalized ranking of topics. Our novel approach follows a selective sampling strategy to perform online model updates based on active learning principles, that closely simulates the task of identifying relevant items from a pool of mostly uninteresting ones. RMFX is particularly suitable for large scale applications and experiments on the 476 million Twitter tweets dataset show that our online approach largely outperforms recommendations based on Twitter's global trend, and it is also able to deliver highly competitive Top-N recommendations faster while using less space than Weighted Regularized Matrix Factorization (WRMF), a state-of-the-art matrix factorization technique for Collaborative Filtering, demonstrating the efficacy of our approach.",
keywords = "Collaborative filtering, Matrix factorization, Online learning, Ranking, Selective sampling, Twitter",
author = "Ernesto Diaz-Aviles and Lucas Drumond and Lars Schmidt-Thieme and Wolfgang Nejdl",
year = "2012",
month = sep,
day = "9",
doi = "10.1145/2365952.2365968",
language = "English",
isbn = "9781450312707",
series = "RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems",
pages = "59--66",
booktitle = "RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems",
note = "6th ACM Conference on Recommender Systems, RecSys 2012 ; Conference date: 09-09-2012 Through 13-09-2012",

}

Download

TY - GEN

T1 - Real-time top-N recommendation in social streams

AU - Diaz-Aviles, Ernesto

AU - Drumond, Lucas

AU - Schmidt-Thieme, Lars

AU - Nejdl, Wolfgang

PY - 2012/9/9

Y1 - 2012/9/9

N2 - The SocialWeb is successfully established, and steadily growing in terms of users, content and services. People generate and consume data in real-time within social networking services, such as Twitter, and increasingly rely upon continuous streams of messages for real-time access to fresh knowledge about current affairs. In this paper, we focus on analyzing social streams in real-time for personalized topic recommen- dation and discovery. We consider collaborative filtering as an online ranking problem and present Stream Ranking Matrix Factorization { RMFX {, which uses a pairwise approach to matrix factorization in order to optimize the personalized ranking of topics. Our novel approach follows a selective sampling strategy to perform online model updates based on active learning principles, that closely simulates the task of identifying relevant items from a pool of mostly uninteresting ones. RMFX is particularly suitable for large scale applications and experiments on the 476 million Twitter tweets dataset show that our online approach largely outperforms recommendations based on Twitter's global trend, and it is also able to deliver highly competitive Top-N recommendations faster while using less space than Weighted Regularized Matrix Factorization (WRMF), a state-of-the-art matrix factorization technique for Collaborative Filtering, demonstrating the efficacy of our approach.

AB - The SocialWeb is successfully established, and steadily growing in terms of users, content and services. People generate and consume data in real-time within social networking services, such as Twitter, and increasingly rely upon continuous streams of messages for real-time access to fresh knowledge about current affairs. In this paper, we focus on analyzing social streams in real-time for personalized topic recommen- dation and discovery. We consider collaborative filtering as an online ranking problem and present Stream Ranking Matrix Factorization { RMFX {, which uses a pairwise approach to matrix factorization in order to optimize the personalized ranking of topics. Our novel approach follows a selective sampling strategy to perform online model updates based on active learning principles, that closely simulates the task of identifying relevant items from a pool of mostly uninteresting ones. RMFX is particularly suitable for large scale applications and experiments on the 476 million Twitter tweets dataset show that our online approach largely outperforms recommendations based on Twitter's global trend, and it is also able to deliver highly competitive Top-N recommendations faster while using less space than Weighted Regularized Matrix Factorization (WRMF), a state-of-the-art matrix factorization technique for Collaborative Filtering, demonstrating the efficacy of our approach.

KW - Collaborative filtering

KW - Matrix factorization

KW - Online learning

KW - Ranking

KW - Selective sampling

KW - Twitter

UR - http://www.scopus.com/inward/record.url?scp=84867345632&partnerID=8YFLogxK

U2 - 10.1145/2365952.2365968

DO - 10.1145/2365952.2365968

M3 - Conference contribution

AN - SCOPUS:84867345632

SN - 9781450312707

T3 - RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems

SP - 59

EP - 66

BT - RecSys'12 - Proceedings of the 6th ACM Conference on Recommender Systems

T2 - 6th ACM Conference on Recommender Systems, RecSys 2012

Y2 - 9 September 2012 through 13 September 2012

ER -

By the same author(s)