Modelling the Substitutability of Discourse Connectives

Proceedings of the 43rd Annual Meeting of the ACL, pages 149–156, Ann Arbor, June 2005. c⃝2005 Association for Computational Linguistics Modelling the substitutability of discourse connectives Ben Hutchinson School of Informatics University of Edinburgh B.Hutchinson@sms.ed.ac.uk Abstract Processing discourse connectives is im- portant for tasks such as discourse parsing and generation. For these tasks, it is use- ful to know which connectives can signal the same coherence relations. This paper presents experiments into modelling the substitutability of discourse connectives. It shows that substitutability effects dis- tributional similarity. A novel variance- based function for comparing probability distributions is found to assist in predict- ing substitutability. 1 Introduction Discourse coherence relations contribute to the meaning of texts, by specifying the relationships be- tween semantic objects such as events and propo- sitions. They also assist in the interpretation of anaphora, verb phrase ellipsis and lexical ambigu- ities (Hobbs, 1985; Kehler, 2002; Asher and Las- carides, 2003). Coherence relations can be implicit, or they can be signalled explicitly through the use of discourse connectives, e.g. because, even though. For a machine to interpret a text, it is impor- tant that it recognises coherence relations, and so as explicit markers discourse connectives are of great assistance (Marcu, 2000). When discourse con- nectives are not present, the task is more difficult. For such cases, unsupervised approaches have been developed for predicting relations, by using sen- tences containing discourse connectives as training data (Marcu and Echihabi, 2002; Lapata and Las- carides, 2004). However the nature of the relation- ship between the coherence relations signalled by discourse connectives and their empirical distribu- tions has to date been poorly understood. In par- ticular, one might wonder whether connectives with similar meanings also have similar distributions. Concerning natural language generation, texts are easier for humans to understand if they are coher- ently structured. Addressing this, a body of research has considered the problems of generating appropri- ate discourse connectives (for example (Moser and Moore, 1995; Grote and Stede, 1998)). One such problem involves choosing which connective to gen- erate, as the mapping between connectives and re- lations is not one-to-one, but rather many-to-many. Siddharthan (2003) considers the task of paraphras- ing a text while preserving its rhetorical relations. Clauses conjoined by but, or and when are sepa- rated to form distinct orthographic sentences, and these conjunctions are replaced by the discourse ad- verbials however, otherwise and then, respectively. The idea underlying Siddharthan’s work is that one connective can be substituted for another while preserving the meaning of a text. Knott (1996) studies the substitutability of discourse connectives, and proposes that substitutability can motivate the- ories of discourse coherence. Knott uses an empiri- cal methodology to determine the substitutability of pairs of connectives. However this methodology is manually intensive, and Knott derives relationships for only about 18% of pairs of connectives. It would thus be useful if substitutability could be predicted automatically. 149 This paper proposes that substitutability can be predicted through statistical analysis of the contexts in which connectives appear. Similar …

Content truncated to 500 words