World Wide Web is a huge repository of information. Retrieving data patterns is facile by using data mining techniques. However identifying the knowledge is tough, tough because the knowledge should be meaningful. Semantics, a branch of linguistics, defines the process of supplying knowledge to the computer system. The underlying idea of semantics is to understand the language model and its correspondence with the meaning associability. Though semantics indicates a crucial ingredient for language processing, the degree of work composition done in this area is minimal. This paper presents an ongoing semantic research problem thereby investigating the theory and rule representation. Probabilistic approach for semantics is demonstrated to address the semantics knowledge representation. The inherit requirement for our system is to have the language syntactically correct. This approach identifies the meaning of the sentence at word-level. The accuracy of the proposed architecture is studied in terms of recall and precision measures. From the experiments conducted, it is clear that the probabilistic model for semantics is able to associate the language model at a preliminary level.

Knowledge processing is the prime area of information retrieval in the current era. However knowledge is subjected to the meaning of discretion in any natural language. Intelligent search in various Natural Languages is required in the huge repository of information available online. Language is the integral part for any form of communication but the language has to be meaningful. Semantics is a field of linguistics that deals with the meaning of the linguistic expressions through discovery of knowledge. In this research paper, the dictionary based approach for semantics is studied and implemented. The dictionary based proposal relies on the formalization of sentence across SVO (Subject-Verb-Object) format. Rule-based classifier helps to define the rules that are checked against the dictionary which contains sequence of Subject, Verbs and Object available in English Language. By looking at the accuracy measures, recall and precision the results obtained by the proposed approach is proven good.