Improving Hypernymy Extraction with Distributional Semantic Classes

Improving Hypernymy Extraction with Distributional Semantic Classes Alexander Panchenko1∗, Dmitry Ustalov2,3∗, Stefano Faralli3, Simone P. Ponzetto3, Chris Biemann1 ∗ these authors contributed equally 1 University of Hamburg, Department of Informatics, Language Technology Group, Germany 2 University of Mannheim, School of Business Informatics and Mathematics, Data and Web Science Group, Germany 3 Ural Federal University, Institute of Natural Sciences and Mathematics, Russia {panchenko,biemann}@informatik.uni-hamburg.de {dmitry,stefano,simone}@informatik.uni-mannheim.de Abstract In this paper, we show how distributionally-induced semantic classes can be helpful for extracting hypernyms. We present methods for inducing sense-aware semantic classes using distributional semantics and using these induced semantic classes for filtering noisy hypernymy relations. Denoising of hypernyms is performed by labeling each semantic class with its hypernyms. On the one hand, this allows us to filter out wrong extractions using the global structure of distributionally similar senses. On the other hand, we infer missing hypernyms via label propagation to cluster terms. We conduct a large-scale crowdsourcing study showing that processing of automatically extracted hypernyms using our approach improves the quality of the hypernymy extraction in terms of both precision and recall. Furthermore, we show the utility of our method in the domain taxonomy induction task, achieving the state-of-the-art results on a SemEval’16 task on taxonomy induction. Keywords: semantic classes, distributional semantics, hypernyms, co-hyponyms, word sense induction 1. Introduction Hypernyms are useful in various applications, such as ques- tion answering (Zhou et al., 2013), query expansion (Gong et al., 2005), and semantic role labelling (Shi and Mihalcea, 2005) as they can help to overcome sparsity of statistical models. Hypernyms are also the building blocks for learn- ing taxonomies from text (Bordea et al., 2016). Consider the following sentence: “This caf´e serves fresh mangosteen juice”. Here the infrequent word “mangosteen” may be poorly represented or even absent in the vocabulary of a sta- tistical model, yet it can be substituted by lexical items with better representations, which carry close meaning, such as its hypernym “fruit” or one of its close co-hyponyms, e.g. “mango”. Currently available approaches to hypernymy extraction fo- cus on the acquisition of individual binary hypernymy rela- tions (Hearst, 1992; Snow et al., 2004; Weeds et al., 2014; Shwartz et al., 2016; Glavaˇs and Ponzetto, 2017). Frequen- cies of the extracted relations usually follow a power-law, with a long tail of noisy extractions containing rare words. We propose a method that performs post-processing of such noisy binary hypernyms using distributional semantics, cf. Figure 1. Namely, we use the observation that distribu- tionally related words are often are co-hyponyms (Wand- macher, 2005; Heylen et al., 2008) and operationalize it to perform filtering of noisy relations by finding dense graphs composed of both hypernyms and co-hyponyms. The contribution of the paper is an unsupervised method for post-processing of noisy hypernymy relations based on clustering of graphs of word senses induced from text. The idea to use distributional semantics to find hypernyms seems natural and has been widely used. However, the ex- isting methods used distributional, yet sense-unaware and local features. We are the first to …

Content truncated to 500 words