Generating Sentences from Different Perspectives

Generating Sentences from Different Perspectives Lee Fedder, The Computer Laboratory, University of Cambridge, Pembroke Street, Cambridge CB2 3QG, England. If@uk.ac.cam.cl Keywords : Generation, Natural Language interfaces. Abstract Certain pairs or groups of sentences appear to be semantically distinct, yet specify the same underlying state of affairs, from different per- spectives. This leads to questions about what that underlying state of affairs might be, and, for generation, how and why the alternative ex- pressions might be produced. This paper looks at how such sentences may be generated in a Natural Language interface to a database sys- tem. Introduction The following sentences would have a different semantics if parsed, yet they seem to specify the same state of affairs at some level of represen- tation. la. I can stay until 5. lb. I must leave by 5. For generation, we ought to be able to pro- duce either. McDonald comments on these sen- tences :- "What mutually known cognitive structure do we recognise from them that would show them to be two sides of the same coin?" (McDonald 1988) This paper describes a language generation system which is designed as the output com- ponent of a database interface, and is capa- ble of producing similar synonymous sentences. The architecture relies on a two level semantic representation: one describes data in the sys- tem's application database, and plays the role of McDonald's "mutually known cognitive struc- ture"; the other describes the semantics of sen- tences of Natural Language, and the primitives correspond to specific entries in the lexicon. In- formation to be communicated is initially ex- pressed in the application level semantics, and is be mapped to the language level semantics as part of the generation process. Alternatives similar to la and lb arise during this mapping, and represent a complexity inherent in language which did not exist in the original data:- they are a property of the description. Application level information is described by linking it with an event or state (from now on the term "event" will cover both these), for which it provides some parameter. Thus, the origin of a flight could be described by saying that the plane "flies from" the origin. The map- ping process exploits a "domain model" which has two parts. The first lays out how non- temporal information is related to domain events. The second describes the temporal character- istics these events using an ontology which is rich enough to capture the temporal semantics of English expressions. Temporal information - 125 - from the application is described by first ex- pressing it in a way that relates it to times in the model, and by then attempting to add it to the description of the event which is currently active. The alternatives arise when more than one event can be used. The temporal ontology is based on a re- cent theory of temporal semantics developed by Moens and Steedman (1988). This allows a modular representation of the semantics of temporal adverbials like "until" and "by", …

Content truncated to 500 words