Semantic Taxonomy Induction from Heterogenous Evidence

By Rion Snow, Dan Jurafsky, Andrew Y. Ng. 2006

551 citations recorded by the source corpus.

We propose a novel algorithm for inducing semantic taxonomies. Previous algorithms for taxonomy induction have typically focused on independent classifiers for discovering new single relationships based on hand-constructed or automatically discovered textual patterns. By contrast, our algorithm flexibly incorporates evidence from multiple classifiers over heterogenous relationships to optimize the entire structure of the taxonomy, using knowledge of a word's coordinate terms to help in determining its hypernyms, and vice versa. We apply our algorithm on the problem of sense-disambiguated noun hyponym acquisition, where we combine the predictions of hypernym and coordinate term classifiers with the knowledge in a preexisting semantic taxonomy (WordNet 2.1). We add 10,000 novel synsets to WordNet 2.1 at 84% precision, a relative error reduction of 70% over a non-joint algorithm using the same component classifiers. Finally, we show that a taxonomy built using our algorithm shows a 23% relative F-score improvement over WordNet 2.1 on an independent testset of hypernym pairs.

Subfields: Artificial Intelligence, Molecular Biology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Molecular Biology

Citations in the corpus

Ordinal Common-sense Inference (2016, background): “Many have pursued automating the process, such as in expanding lexical hierarchies (Hearst, 1992; Snow et al., 2006), constructing inference patterns (Lin and Pantel, 2001; Berant et al.”
Learning Knowledge Graphs for Question Answering through Conversational Dialog (2015, methodology): “For example, Snow et al (2006) use Wordnet as training data for ontology induction.”
NaturalLI: Natural Logic Inference for Common Sense Reasoning (2014, methodology): “For instance, Snow et al. (2006) present an approach to enriching the WordNet taxonomy; Tandon et al. (2011) extend ConceptNet with new facts; Soderland et al. (2010) use ReVerb extractions to enrich a domain-specific ontology.”
YAGO: A Core of Semantic Knowledge Unifying WordNet and Wikipedia (2007, background): “State-ofthe-art taxonomy induction as described in [23] achieves an accuracy of 84%.”
Deriving a Large-Scale Taxonomy from Wikipedia (2007, background): “Using PageRank filtering together with paths including only isa relations yields results which are close to the ones obtained by using WordNet.”
Large-Scale, Diverse, Paraphrastic Bitexts via Sampling and Clustering (2019, background): “There have been efforts to augment such resources with automatic meth-ods (Snow et al., 2006; Pavlick et al., 2015b), but they are still conﬁned to lexical level and sometimes require the use of other paraphrastic resources (Pavlick et al., 2015b).”
Learning Entailment Relations by Global Graph Structure Optimization (2012, methodology): “(2) Greedy-Global : An algorithm that looks for the maximum a posteriori graphs but only employs the greedy optimization procedure as described by Snow, Jurafsky, and Ng (2006).”
Learning Semantic Hierarchies via Word Embeddings (2014, methodology): “Snow et al. (2006) provides a global optimization scheme for extending WordNet, which is different from the above-mentioned pairwise relationships identification methods.”
Vector Space Models of Word Meaning and Phrase Meaning: A Survey (2012, background): “…et al. 2008), word sense induction (Schütze 1998), ontology learning (Hindle 1990; Lin 1998; Ravichandran et al. 2005; Gorman and Curran 2006; Snow et al. 2006), determining predominant word sense (McCarthy et al. 2004, Padó and Lapata 2007), predicting similarity of semantic relations…”
Hierarchical Semi-supervised Classification with Incomplete Class Hierarchies (2016, background): “Snow et al. [22] discover new WordNet synsets by using evidence from multiple sources, however their approach is focused on discovering new isA relations, and not meant for building hierarchical classi-ﬁers for the learnt hierarchy.”
YAGO: A Large Ontology from Wikipedia and WordNet (2008, methodology): “These approaches use information extraction technologies that include pattern matching, natural-language parsing, and statistical learning [49,25,11,1,46,41,18].”
Random Walk Inference and Learning in A Large Scale Knowledge Base (2011, background): “Although there is a great deal of recent research on extracting knowledge from text (Agichtein and Gravano, 2000; Etzioni et al., 2005; Snow et al., 2006; Pantel and Pennacchiotti, 2006; Banko et al., 2007; Yates et al., 2007), much less progress has been made on the problem of drawing reliable infe”
Evaluating the Inferential Utility of Lexical-Semantic Resources (2009, methodology): “Snow ( Snow 30 k ) : Snow et al. (2006) presented a probabilistic model for taxonomy induction which considers as features paths in parse trees between related taxonomy nodes.”
Recovering Semantics of Tables on the Web (2011, background): “When associated with a class label, the sets of instances may be organized as ﬂat sets or hierarchically, relative to existing hierarchies such as WordNet [25, 26] or the category network within Wikipedia [24, 31].”
Joint learning of ontology and semantic parser from text (2016, unknown): “Also, many other approaches focus only on one aspect of knowledge extraction, like taxonomy extraction [26,31] or relation extraction [7,39].”
BabelNet: Building a Very Large Multilingual Semantic Network (2010, background): “We also intend to link missing concepts in WordNet, by establishing their most likely hypernyms – e.g., `a la Snow et al. (2006).”