Ontology-based Prediction of Compound Relations : A Study Based on SUMO

Ontology-based Prediction of Compound Relations — A study based on SUMO Jia-Fei Hong Institute of Linguistics Academia Sinica jiafei@gate.sinica.edu.tw Xiang-Bing Li Institute of Information Science, Academia Sinica dreamer@hp.iis.sinica.edu.tw Chu-Ren Huang Institute of Linguistics Academia Sinica churen@gate.sinica.edu.tw Abstract This paper explores the interaction between conceptual structure and morpho-syntax. In particular, we show that ontology-based conceptual classification can be used to predict internal relations in compounds. We propose an ontology-based approach to predict the semantic relation between the two component words in Mandarin VV compounds. A Mandarin VV compound is classified according to the eventive relation between the two simplex verbs. These relations specify how the eventive meanings of the two simplex verbs combine to form the meaning of the compound. The three types of eventive relations that we deal with in this paper are: coordinate, modificational, and resultative. Since the way in which two events combine with each other depends upon their event types, we hypothesize that the eventive relations can be predicted by the conceptual classified event types of the two simplex verbs. An approach of ontology-based prediction is proposed based on this hypothesis. The assignment of ontology classification for each simplex verb is based on SUMO and Sinica BOW. The correlation between the ontology class of each verb position and each eventive type is trained and scored based on a manually tagged lexical database. We encode the ontology information of each VV compound in a 3-tuple based on these correlation scores. This 3-tuple is represented as a three-dimensional vector and used to predict the eventive type of new VV compounds. Our classification experiment on unknown VV compounds yields good recall and precision. 1. Introduction This paper explores the interaction between conceptual structure and morpho-syntax. In particular, we demonstrate that ontology-based conceptual classification can predict the internal relations in compounds. Even though the interaction between lexical semantics and syntax has been a PACLIC 18, December 8th-10th, 2004, Waseda University, Tokyo - 151 - fertile ground for research in both theoretical and computational linguistics, there has been limited work on the interaction on semantics and morphology. This is somewhat unexpected since both lexical semantics and morphology are lexical operations. In Chinese, for instance, study on the formation of compound verbs has so far focused on the grammatical categories and grammatical functions of the simplex words being combined. However, it is clear that the lexical meaning of the simplex words must be preserved in compound formation. In other words, simplex words are building blocks of compounds that necessarily contain their grammatical information, including lexical meaning. Given the existence of the lexical meanings in the component words, an interesting theoretical issue is how they combine to form the meaning of the compound and to examine if the combination is governed by structural or conceptual principles? In this paper, we deal with the internal relations of the VV compounds. Since the two components are the same lexical category, the internal relation is more likely to be conceptual. Our goal is to show that the …

Content truncated to 500 words