Text classification and named entities for new event detection
G. Kumaran, and J. Allan. SIGIR '04: Proceedings of the 27th annual international ACM SIGIR conference on Research and development in information retrieval, page 297--304. New York, NY, USA, ACM Press, (2004)
DOI: 10.1145/1008992.1009044
Abstract
New Event Detection is a challenging task that still offers scope for great improvement after years of effort. In this paper we show how performance on New Event Detection (NED) can be improved by the use of text classification techniques as well as by using named entities in a new way. We explore modifications to the document representation in a vector space-based NED system. We also show that addressing named entities preferentially is useful only in certain situations. A combination of all the above results in a multi-stage NED system that performs much better than baseline single-stage NED systems.
%0 Conference Paper
%1 citeulike:1219863
%A Kumaran, Giridhar
%A Allan, James
%B SIGIR '04: Proceedings of the 27th annual international ACM SIGIR conference on Research and development in information retrieval
%C New York, NY, USA
%D 2004
%I ACM Press
%K named-entity news
%P 297--304
%R 10.1145/1008992.1009044
%T Text classification and named entities for new event detection
%U http://dx.doi.org/10.1145/1008992.1009044
%X New Event Detection is a challenging task that still offers scope for great improvement after years of effort. In this paper we show how performance on New Event Detection (NED) can be improved by the use of text classification techniques as well as by using named entities in a new way. We explore modifications to the document representation in a vector space-based NED system. We also show that addressing named entities preferentially is useful only in certain situations. A combination of all the above results in a multi-stage NED system that performs much better than baseline single-stage NED systems.
%@ 1581138814
@inproceedings{citeulike:1219863,
abstract = {New Event Detection is a challenging task that still offers scope for great improvement after years of effort. In this paper we show how performance on New Event Detection (NED) can be improved by the use of text classification techniques as well as by using named entities in a new way. We explore modifications to the document representation in a vector space-based NED system. We also show that addressing named entities preferentially is useful only in certain situations. A combination of all the above results in a multi-stage NED system that performs much better than baseline single-stage NED systems.},
added-at = {2009-07-01T11:12:30.000+0200},
address = {New York, NY, USA},
author = {Kumaran, Giridhar and Allan, James},
biburl = {https://www.bibsonomy.org/bibtex/2042c7a381dbcf1aa507b4fa9bf558187/brusilovsky},
booktitle = {SIGIR '04: Proceedings of the 27th annual international ACM SIGIR conference on Research and development in information retrieval},
citeulike-article-id = {1219863},
doi = {10.1145/1008992.1009044},
interhash = {70772f5cbd8aab416d12db6303928a2f},
intrahash = {042c7a381dbcf1aa507b4fa9bf558187},
isbn = {1581138814},
keywords = {named-entity news},
pages = {297--304},
posted-at = {2008-09-18 19:59:59},
priority = {2},
publisher = {ACM Press},
timestamp = {2009-07-01T11:12:35.000+0200},
title = {Text classification and named entities for new event detection},
url = {http://dx.doi.org/10.1145/1008992.1009044},
year = 2004
}