Abstract

Information Extraction (IE) approaches currently assume that a template exists which sufficiently defines the requirements of the task. Substantial human effort is required to generate these basic templates and to provide a development corpus. In the two principal IE competitions, the Message Understanding Conference (MUC) and Tipster, the templates were constructed directly from the experience of analysts. This manual approach cannot always be assumed. This proposal concerns the automatic construction of MUC-style templates, substantially reducing the human effort required. The approach will carry out a corpus-based analysis of task-relevant documents, identifying and analysing the interaction between the fundamental elements. A resource which defines semantic relationships will be necessary to identify and categorise these fundamental elements. This application is of particular interest to researchers in the field of IE and automatic abstracting. 1 1.

Citations

1654	B.: Building a large annotated corpus of english: The Penn treebank - Marcus, Marcinkiewicz, et al. - 1993
1302	WordNet: an online lexical database - Miller, Beckwith, et al. - 1990
642	Accurate methods for the statistics of surprise and coincidence - Dunning - 1993
183	Automatically Constructing a Dictionary for Information Extraction Tasks - Riloff - 1993
136	Crystal: Inducing a conceptual dictionary - Soderland, Fisher, et al. - 1995
132	Some Advances in Rulebased Part of Speech Tagging - Brill - 1994
131	Using statistics in lexical analysis - Church, Gale, et al. - 1991
128	C.: The Unified Medical Language System - Lindberg, Humphrey, et al.
106	Longman dictionary of contemporary English - Procter - 1978
104	Evaluating natural language processing systems - Jones, K, et al. - 1996
92	The Analysis of Cross-Classified Categorical Data - Fienberg - 1977
72	The NYU system for MUC-6 or Where’s the Syntax - Grishman - 1995
67	MITRE: description of the alembic system used for MUC-6 - Aberdeen, Day, et al. - 1995
63	Description of the UMass System as Used for MUC-6 - Fisher, Soderland, et al. - 1995
58	Description of the SRA System as Used for MUC-6 - Krupka - 1995
41	A new method of n-gram statistics for large number of n and automatic extraction of words and phrases from large text data of Japanese - Nagao, Mori - 1994
35	University of Sheffield: Description of the LaSIE System as Used for MUC-6 - Gaizauskas, Humphreys, et al. - 1995
31	Acquisition of semantic patterns for information extraction from corpora - Kim, Moldovan - 1993
23	Automatically Acquiring Conceptual Patterns without an Automated Corpus - Riloff, Shoen - 1995
11	Tipster Phase II Architecture Design Document, Version 1.52 - Grishman - 1995
7	Principles of template design - Hobbs, Israel - 1994
7	Towards a Workbench for Acquisition of Domain Knowledge from Natural Language - Mikheev, Finch - 1995
5	Automatically Acquiring Conceptual Patterns Without an Annotated Corpus - Riloff, Shoen - 1995
5	The development of an application independent lexicon: LEXBASE - Stein, Lin, et al. - 1993
4	G.: Introduction to Mathematical Statistics, Fourth ed - Hoel - 1971
4	Template Design for Information Extraction - Onyshkevych - 1993
2	N-gram cluster identification during empirical knowledge representation generation - Collier - 1994
2	Financial information extraction using predefined and user-definable templates in the LOLITA system - Costantino, Morgan, et al. - 1996
1	Appendix B: MUC-6 test scores - 1996a
1	Four scorers and seven years ago - Chinchor - 1996
1	An introduction to lexical resources at Sheffield - Stevenson, Azzam, et al. - 1996