Font Size: a A A

Construction Of Domain Dictionary And Its Application In Grammar Analysis

Posted on:2012-03-11Degree:MasterType:Thesis
Country:ChinaCandidate:L JiangFull Text:PDF
GTID:2218330338463055Subject:Computer software and theory
Abstract/Summary:PDF Full Text Request
The concept of domain dictionary has gradually entered in our lives. The traditionalgeneral-dictionary is quite ripe, but it's inadequate in application of specific domain. We not onlyneed to mark the syntax, but also try to establish a relatively perfect semantic network system.Because of rapid development in every area nowadays, new vocabulary, concepts and relations haveemerged continually. If we still stuck in the traditional vocabulary of analysis, it fell far short of ourneeds. But the emergence of the domain dictionary is able to largely solve this problem. We cancollect the latest concepts and relations between them by constructing domain dictionaries, so that itmakes our research on specific domain more convenient. In practice, we often need detailedanalysis to concepts in specific domain, such as research on semantic relations and someapplications base on the relation of concepts.This paper discusses the relationship between concepts, such as hyponymy,synonyms/antonym, meronym/holonym, host/property and entailment relationship and so on.Meanwhile, according to the speech and classification of the concepts, we use JSON as a dataformat to construct the dictionary. Additionally, we can reduce the time in finding a concept in adictionary by designing coding rules and index table. So we put forward the method of constructingthe domain dictionary base on relationship of concepts. Then, we apply the method of constructingdomain dictionary discussed before in programmed grammar rules, in order to study and developprogrammed set of grammar rules that we express it with appropriate grammar symbols. Weanalyze the relationship between these symbols, including the entailment relation, constitutiverelation, first and follow relation to construct grammar rule dictionary. According to the existentrule dictionary, we do grammar analysis base on rule tree (bottom-up base on rule tree). With thematching between rule tree and symbol string, and judging whether current input char and thesymbol at the top of the symbol stack can form follow relationship, we establish shift/reduce criteriato do grammar analysis. At last, we conduct an experiment to analyze BUBRT method. Weintroduce the runtime environment of the experiment and functional modules to analyze the relationof these modules. Finally, we prove that BUBRT method is feasible according to the result of theexperiment.
Keywords/Search Tags:domain dictionary, relationship of concepts, JSON, grammar rules, grammaranalysis, BUBRT method
PDF Full Text Request
Related items