Supervised and unsupervised approaches to measuring usage similarity
Milton King, Paul F. Cook · 2017
Usage similarity (USim) is an approach to determining word meaning in context that does not rely on a sense inventory.Instead, pairs of usages of a target lemma are rated on a scale.In this paper we propose unsupervised approaches to USim based on embeddings for words, contexts, and sentences, and achieve state-of-the-art results over two USim datasets.We further consider supervised approaches to USim, and find that although they outperform unsupervised approaches, they are unable to generalize to lemmas that are unseen in the training data. Usage similarityWord senses are not discrete.In many cases, for a given instance of a word, multiple senses from a sense inventory are applicable, and to varying degrees (Erk et al., 2009).For example, consider the usage of wait in the following sentence taken from Jurgens and Klapaftis (2013):