Automatic disambiguation of English puns

Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing, pages 719–729, Beijing, China, July 26-31, 2015. c⃝2015 Association for Computational Linguistics Automatic disambiguation of English puns Tristan Miller and Iryna Gurevych Ubiquitous Knowledge Processing Lab (UKP-TUDA) Department of Computer Science, Technische Universit¨at Darmstadt https://www.ukp.tu-darmstadt.de/ Abstract Traditional approaches to word sense dis- ambiguation (WSD) rest on the assump- tion that there exists a single, unambigu- ous communicative intention underlying every word in a document. However, writ- ers sometimes intend for a word to be in- terpreted as simultaneously carrying mul- tiple distinct meanings. This deliberate use of lexical ambiguity—i.e., punning— is a particularly common source of humour. In this paper we describe how traditional, language-agnostic WSD approaches can be adapted to “disambiguate” puns, or rather to identify their double meanings. We eval- uate several such approaches on a manually sense-annotated collection of English puns and observe performance exceeding that of some knowledge-based and supervised baselines. 1 Introduction Word sense disambiguation, or WSD, is the task of identifying a word’s meaning in context. No matter whether it is performed by a human or a machine, WSD usually rests on the assumption that there is a single unambiguous communicative intention underlying each word in the document.1 However, there exists a class of language constructs known 1Under this assumption, lexical ambiguity arises due to there being a plurality of words with the same surface form but different meanings, and the task of the interpreter is to select correctly among them. An alternative view is that each word is a single lexical entry whose specific meaning is un- derspecified until it is activated by the context (Ludlow, 1996). In the case of systematically polysemous terms (i.e., words that have several related senses shared in a systematic way by a group of similar words), it may not be necessary to disam- biguate them at all in order to interpret the communication (Buitelaar, 2000). While there has been some research in mod- elling intentional lexical-semantic underspecification (Jurgens, 2014), it is intended for closely related senses such as those of systematically polysemous terms, not those of coarser-grained homonyms which are the subject of this paper. as paronomasia and syllepsis, or more generally as puns, in which homonymic (i.e., coarse-grained) lexical-semantic ambiguity is a deliberate effect of the communication act. That is, the writer intends for a certain word or other lexical item to be in- terpreted as simultaneously carrying two or more separate meanings, or alternatively for it to be un- clear which meaning is the intended one. There are a variety of motivations writers have for employing such constructions, and in turn for why such uses are worthy of scholarly investigation. Perhaps surprisingly, this sort of intentional lex- ical ambiguity has attracted little attention in the fields of computational linguistics and natural lan- guage processing. What little research has been done is confined largely to computational mecha- nisms for pun generation (in the context …

Content truncated to 500 words