Unsupervised information extraction from unstructured, ungrammatical data sources on the World Wide Web Export

@article{citeulike:2641210, abstract = {Abstract\ \ Information extraction from unstructured, ungrammatical data such as classified listings is difficult because traditional structural and grammatical extraction methods do not apply. Previous work has exploited reference sets to aid such extraction, but it did so using supervised machine learning. In this paper, we present an unsupervised approach that both selects the relevant reference set(s) automatically and then uses it for unsupervised extraction. We validate our approach with experimental results that show our unsupervised extraction is competitive with supervised machine learning approaches, including the previous supervised approach that exploits reference sets.}, author = {Michelson, Matthew and Knoblock, Craig}, citeulike-article-id = {2641210}, citeulike-linkout-0 = {http://dx.doi.org/10.1007/s10032-007-0052-2}, doi = {10.1007/s10032-007-0052-2}, journal = {International Journal on Document Analysis and Recognition}, keywords = {ir, spelling\_variation, string\_comparison}, month = {December}, number = {3}, pages = {211--226}, posted-at = {2008-04-08 11:24:44}, priority = {2}, title = {Unsupervised information extraction from unstructured, ungrammatical data sources on the World Wide Web}, url = {http://dx.doi.org/10.1007/s10032-007-0052-2}, volume = {10}, year = {2007} }

TY - JOUR ID - citeulike:2641210 L3 - citeulike-article-id:2641210 N2 - Abstract  Information extraction from unstructured, ungrammatical data such as classified listings is difficult because traditional structural and grammatical extraction methods do not apply. Previous work has exploited reference sets to aid such extraction, but it did so using supervised machine learning. In this paper, we present an unsupervised approach that both selects the relevant reference set(s) automatically and then uses it for unsupervised extraction. We validate our approach with experimental results that show our unsupervised extraction is competitive with supervised machine learning approaches, including the previous supervised approach that exploits reference sets. IS - 3 JF - International Journal on Document Analysis and Recognition EP - 226 TI - Unsupervised information extraction from unstructured, ungrammatical data sources on the World Wide Web VL - 10 SP - 211 KW - ir KW - spelling_variation KW - string_comparison AU - Michelson, Matthew AU - Knoblock, Craig PY - 2007/12// UR - http://dx.doi.org/10.1007/s10032-007-0052-2 ER -