期刊名称：BVICAM's International Journal of Information Technology
印刷版ISSN：0973-5658
出版年度：2013
卷号：5
期号：1
语种：English
出版社：Bharati Vidyapeeth's Institute of Computer Applications and Management
摘要：Knowledge representation is base for expressing semantic content of input in intelligent information retrieval systems. Identification of semantic requires processing of input language at various levels. To make system understand text or speech is a challenging task as it involves extracting semantics of the language which itself is a complex problem. At the same time languages posses with multiple ambiguities and uncertainty which needs to be resolved at various phases of language processing. Level of understandability depends upon the grammar, syntactic and semantic representation of the language and methods employed for these analysis. Processing depends on the type of language, grammar of the language, ambiguities present and size of corpus available. Order free language posses different features as compared to rigid order language. Most of the Indian languages are order free; hence mechanism for such language needs to be formulated. One of the ancient Indian Sanskrit grammarians, pAninI has defined grammar of Sanskrit language in such a way that it is suitable for computational analysis. Six main semantic class identified under this theory is a baseline model for knowledge representation. This paper exploits the features of the language, applicability of rules and resolving ambiguities using neural network model. A hybrid model incorporating the features of rules based and neural network the is designed and implemented for pAninI based semantic analysis, generating case frames as output.
