Combining concept hierarchies and statistical topic models Export

@inproceedings{citeulike:4066273, abstract = {Statistical topic models provide a general data-driven framework for automated discovery of high-level knowledge from large collections of text documents. While topic models can potentially discover a broad range of themes in a data set, the interpretability of the learned topics is not always ideal. Human-defined concepts, on the other hand, tend to be semantically richer due to careful selection of words to define concepts but they tend not to cover the themes in a data set exhaustively. In this paper, we propose a probabilistic framework to combine a hierarchy of human-defined semantic concepts with statistical topic models to seek the best of both worlds. Experimental results using two different sources of concept hierarchies and two collections of text documents indicate that this combination leads to systematic improvements in the quality of the associated language models as well as enabling new techniques for inferring and visualizing the semantics of a document.}, address = {New York, NY, USA}, author = {Chemudugunta, Chaitanya and Smyth, Padhraic and Steyvers, Mark}, booktitle = {CIKM '08: Proceeding of the 17th ACM conference on Information and knowledge management}, citeulike-article-id = {4066273}, citeulike-linkout-0 = {http://portal.acm.org/citation.cfm?id=1458082.1458337}, citeulike-linkout-1 = {http://dx.doi.org/10.1145/1458082.1458337}, doi = {10.1145/1458082.1458337}, isbn = {978-1-59593-991-3}, keywords = {u24}, location = {Napa Valley, California, USA}, pages = {1469--1470}, posted-at = {2009-06-11 23:17:12}, priority = {2}, publisher = {ACM}, title = {Combining concept hierarchies and statistical topic models}, url = {http://dx.doi.org/10.1145/1458082.1458337}, year = {2008} }

TY - CONF ID - citeulike:4066273 L3 - citeulike-article-id:4066273 N2 - Statistical topic models provide a general data-driven framework for automated discovery of high-level knowledge from large collections of text documents. While topic models can potentially discover a broad range of themes in a data set, the interpretability of the learned topics is not always ideal. Human-defined concepts, on the other hand, tend to be semantically richer due to careful selection of words to define concepts but they tend not to cover the themes in a data set exhaustively. In this paper, we propose a probabilistic framework to combine a hierarchy of human-defined semantic concepts with statistical topic models to seek the best of both worlds. Experimental results using two different sources of concept hierarchies and two collections of text documents indicate that this combination leads to systematic improvements in the quality of the associated language models as well as enabling new techniques for inferring and visualizing the semantics of a document. CY - Napa Valley, California, USA SN - 978-1-59593-991-3 AD - New York, NY, USA EP - 1470 TI - Combining concept hierarchies and statistical topic models PB - ACM T2 - CIKM '08: Proceeding of the 17th ACM conference on Information and knowledge management SP - 1469 KW - u24 AU - Chemudugunta, Chaitanya AU - Smyth, Padhraic AU - Steyvers, Mark PY - 2008/// UR - http://dx.doi.org/10.1145/1458082.1458337 ER -