Maintaining consistency and plausibility in integrated natural language understanding

In this paper, we present an inference mechanism called the integrated parsing engine which provides a uniform abductive inference mech-anism for natu~al language understmlding. It can (1) make phmsibh, assmnptiox~s, (2) rea~. son with mlfltiple alternatives, (3) switch the search process to the maximally plausible alter.., native, (4) detect contradict ion attd tame COlt -~ clutions which depend on inconsistent a,'~sumptions, and (5) upda te plausibility factor of each befief based on new obsexvations. We demonstrafe that a .natural language unders tanding system using the integrated parsing engine as a subsystem can pursue a guided search for most t)lausible interpretat ion by making use of syntax, semantics, and contextual information. 1 I n t r o d u c t i o n Natural language unders tanding involves lots of hard issues such as various types of ambiguities, indeterrtfinacies caused by ellipses or fragmental utterances, or ill-formedness. Being confronted with these difl]culties, it does not seen, reason-able t o s e e k for a method of logically deducing the spe~ker's intexMed meaning or p]an from utterances. Insteaxt, it is much more natura l to characterize na tura l language unders tandhlg a~q an abduct ive process of exploring most plausb ble interpretat ion which can ext)lain given utt el'a/Ices. In this paper, we present an abductive in-. ference mechanism, called the integrMcd parsing engine, for na tura l lmlguage lmdcrst;mding. The integrated pea'sing engine is ~ble to: make plausible assunrptions at z,:pproprlate t ime reason with mult iple alternatives based o~ ditferent sets of a~ss~m~ptions switch the sem'ch process to the maximally plausible alternative , detect contradiction resulting from inco~v sistent ~ sumpf ions and eliminate ~fil conclutions which depends ,',m these assump~ tions * upda te plausibility factor of each belief based on new observations. Thus, the integrated parsing engine is generM enough to carry out hngulstic and nonlingulso. tic inferences in a uniform manner , by drawing information from various sources: syntax, seo mantle, discourse, pragmatics, or real world. In the remainder of this paper, we first de-~ scribe mechanisms for mainta ining consistency and plausibility. We then show how these two mechauisms interact to guide the inference pro.. tess. Finally, we use an implemented exam= ple to demonstra te how the integrated parsing engine is used to interpret sentences by taking contextual factors into account. 2 M a i n t a i n i n g Cons~si, e:c~cy The CME (Consistency Maintenm~ce Engixtc) is a component of the ht tegrated pa:rsmg (mghte