%0 Conference Paper %1 1014137 %A Popescul, Alexandrin %A Ungar, Lyle H. %B KDD '04: Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining %C New York, NY, USA %D 2004 %I ACM %K learning relational %P 665--670 %R 10.1145/1014052.1014137 %T Cluster-based concept invention for statistical relational learning %U http://portal.acm.org/citation.cfm?id=1014052.1014137&coll=portal&dl=ACM&CFID=83114244&CFTOKEN=43516963 %X We use clustering to derive new relations which augment database schema used in automatic generation of predictive features in statistical relational learning. Entities derived from clusters increase the expressivity of feature spaces by creating new first-class concepts which contribute to the creation of new features. For example, in CiteSeer, papers can be clustered based on words or citations giving "topics", and authors can be clustered based on documents they co-author giving "communities". Such cluster-derived concepts become part of more complex feature expressions. Out of the large number of generated features, those which improve predictive accuracy are kept in the model, as decided by statistical feature selection criteria. We present results demonstrating improved accuracy on two tasks, venue prediction and link prediction, using CiteSeer data. %@ 1-58113-888-1

@inproceedings{1014137, abstract = {We use clustering to derive new relations which augment database schema used in automatic generation of predictive features in statistical relational learning. Entities derived from clusters increase the expressivity of feature spaces by creating new first-class concepts which contribute to the creation of new features. For example, in CiteSeer, papers can be clustered based on words or citations giving "topics", and authors can be clustered based on documents they co-author giving "communities". Such cluster-derived concepts become part of more complex feature expressions. Out of the large number of generated features, those which improve predictive accuracy are kept in the model, as decided by statistical feature selection criteria. We present results demonstrating improved accuracy on two tasks, venue prediction and link prediction, using CiteSeer data.}, added-at = {2010-07-17T13:19:28.000+0200}, address = {New York, NY, USA}, author = {Popescul, Alexandrin and Ungar, Lyle H.}, biburl = {https://www.bibsonomy.org/bibtex/20b9bb78ad6d64bd169b0a79d0413fa65/lst}, booktitle = {KDD '04: Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining}, description = {Cluster-based concept invention for statistical relational learning}, doi = {10.1145/1014052.1014137}, interhash = {4d975e7fac7f5edb80027147f678b1b4}, intrahash = {0b9bb78ad6d64bd169b0a79d0413fa65}, isbn = {1-58113-888-1}, keywords = {learning relational}, location = {Seattle, WA, USA}, pages = {665--670}, publisher = {ACM}, timestamp = {2010-07-17T13:19:29.000+0200}, title = {Cluster-based concept invention for statistical relational learning}, url = {http://portal.acm.org/citation.cfm?id=1014052.1014137&coll=portal&dl=ACM&CFID=83114244&CFTOKEN=43516963}, year = 2004 }