Abstract

In this paper, we present Espresso, a weakly-supervised, general-purpose, and accurate algorithm for harvesting semantic relations. The main contributions are: i) a method for exploiting generic patterns by filtering incorrect instances using the Web; and ii) a principled measure of pattern and instance reliability enabling the filtering algorithm. We present an empirical comparison of Espresso with various state of the art systems, on different size and genre corpora, on extracting various general and specific relations. Experimental results show that our exploitation of generic patterns substantially increases system recall with small effect on overall precision. 1

Citations

6514	Elements of Information Theory - Cover, Thomas - 1991
1996	WordNet: An Electronic Lexical Database - Fellbaum, editor - 1998
771	Nonparametric Statistics for Behavioral Sciences - Siegel, Castellan - 1988
673	Automatic Acquisition of Hyponyms from Large Text Corpora - Hearst - 1992
207	Learning surface text patterns for a question answering system - Ravichandran, Hovy - 2002
205	Unsupervised named-entity extraction from the web: an experimental study - Etzioni, Cafarella, et al. - 2005
201	S: Technical terminology: some linguistic properties and an algorithm for identification in text. Natural Language Engineering - Justeson, Katz - 1995
198	Noun classification from predicate-argument structures - Hindle - 1990
178	Finding parts in very large corpora - Berland, Charniak - 1999
95	A Corpus-Based Approach for Building Semantic Lexicons - Riloff, Shepherd - 1997
71	A probabilistic model of redundancy in information extraction - Downey, Etzioni, et al. - 2005
69	The automated acquisition of topic signatures for text summarization - Lin, Hovy - 2000
56	MixedInitiative Development of Language Processing Systems - Day, Aberdeen, et al. - 1997
47	Concept discovery from text - Lin, Pantel - 2002
47	D.: Automatically labeling semantic classes - Pantel, Ravichandran - 2004
40	Towards terascale knowledge acquisition - Pantel, Ravichandran, et al. - 2004
34	The Informative Role of WordNet in Open-Domain Question Answering - Pasca, Harabagiu - 2001
32	Automatic discovery of part-whole relations - Girju, Badulescu, et al. - 2006
31	Automatic Acquisition of a Hypernym-Labeled Noun Hierarchy from Text - Caraballo - 1999
23	Fine-grained proper noun ontologies for question answering - Mann - 2002
17	The distributional inclusion hypotheses and lexical entailment - Geffet, Dagan - 2005