%0 Conference Paper %1 conf/www/XiangNR10 %A Xiang, Rongjing %A Neville, Jennifer %A Rogati, Monica %B Proceedings of the 19th international conference on World wide web %C New York, NY, USA %D 2010 %E Rappa, Michael %E Jones, Paul %E Freire, Juliana %E Chakrabarti, Soumen %I ACM %K relationshipmining %P 981--990 %R 10.1145/1772690.1772790 %T Modeling relationship strength in online social networks %U http://doi.acm.org/10.1145/1772690.1772790 %X Previous work analyzing social networks has mainly focused on binary friendship relations. However, in online social networks the low cost of link formation can lead to networks with heterogeneous relationship strengths (e.g., acquaintances and best friends mixed together). In this case, the binary friendship indicator provides only a coarse representation of relationship information. In this work, we develop an unsupervised model to estimate relationship strength from interaction activity (e.g., communication, tagging) and user similarity. More specifically, we formulate a link-based latent variable model, along with a coordinate ascent optimization procedure for the inference. We evaluate our approach on real-world data from Facebook and LinkedIn, showing that the estimated link weights result in higher autocorrelation and lead to improved classification accuracy. %@ 978-1-60558-799-8

@inproceedings{conf/www/XiangNR10, abstract = {Previous work analyzing social networks has mainly focused on binary friendship relations. However, in online social networks the low cost of link formation can lead to networks with heterogeneous relationship strengths (e.g., acquaintances and best friends mixed together). In this case, the binary friendship indicator provides only a coarse representation of relationship information. In this work, we develop an unsupervised model to estimate relationship strength from interaction activity (e.g., communication, tagging) and user similarity. More specifically, we formulate a link-based latent variable model, along with a coordinate ascent optimization procedure for the inference. We evaluate our approach on real-world data from Facebook and LinkedIn, showing that the estimated link weights result in higher autocorrelation and lead to improved classification accuracy.}, acmid = {1772790}, added-at = {2015-05-10T10:56:52.000+0200}, address = {New York, NY, USA}, author = {Xiang, Rongjing and Neville, Jennifer and Rogati, Monica}, biburl = {https://www.bibsonomy.org/bibtex/2d2c2579a5ac8e657957b5ca2ebf88a3a/jiangxiluning}, booktitle = {Proceedings of the 19th international conference on World wide web}, doi = {10.1145/1772690.1772790}, editor = {Rappa, Michael and Jones, Paul and Freire, Juliana and Chakrabarti, Soumen}, interhash = {8b1a72c3e48b8327d0494742eb6a8a43}, intrahash = {d2c2579a5ac8e657957b5ca2ebf88a3a}, isbn = {978-1-60558-799-8}, keywords = {relationshipmining}, location = {Raleigh, North Carolina, USA}, numpages = {10}, pages = {981--990}, publisher = {ACM}, series = {WWW '10}, timestamp = {2015-05-10T10:56:52.000+0200}, title = {Modeling relationship strength in online social networks}, url = {http://doi.acm.org/10.1145/1772690.1772790}, year = 2010 }