Q63 : Design and implementation of a lexical dataset named Fuzznet considering fuzzy theory for statistical machine translation
Thesis > Central Library of Shahrood University > Computer Engineering > MSc > 2015
Authors:
Abstarct: By rapid growth of artificial intelligence science, new methods in this field have better performance and fewer errors. Natural language processing is a topic in artificial intelligence. So many articles are written about text mining, most of which are aimed at text classification and some clustering approaches have used fuzzy methods. One of tasks in the field of natural language processing and text mining is creation of lexical databaxse WordNet. Lexical databaxse WordNet is a set of words in synonym sets by which knowledge baxse is created. The databaxse was created by Princeton university and also developed in other languages. FarsNet as a model of WordNet in Persian is implemented by laboratory of natural language processing in Shahid Beheshti university. FarsNet deals with different semantic categories for every word and in fact is an ontology. FarsNet is an introduction to Persian knowledge baxse.
In this thesis modified lexical databaxse named FuzzNet is created by means of lexical databaxse FarsNet. In fact Fuzznet is a combination of FarsNet and fuzzy theory. As regards relation between a word with respected meanings is crisp in FarsNet, the relation is defined fuzzy instead of crisp, That is each word is related to its meaning by fuzzy values. Membership value of a meaning for a word is higher if the meaning is closer to mind. The fuzzy membership value is helpful in translation as that meaning with higher membership value for the word is selected and translation for the meaning is done.
Keywords:
#fuzzy #FarsNet #WordNet #translation #databaxse Link
Keeping place: Central Library of Shahrood University
Visitor:
Keeping place: Central Library of Shahrood University
Visitor: