UNL Based Semantic Search Engine and its Strategies with Interactive ANalyzer

Abstract

A semantic based approach is better in many ways than the keyword based searches. As the wide amount of information on the internet is exceling daily, the keyword based approach in many commercial search engines is failing to retrieve correct information without getting the knowledge and need of user query. Since the keyword based search could search for the particular language having that keyword, thus many commercial search engine is one language oriented. The most challenging task is to design a system for multi-lingual machine translation environment, where large number of languages are to be translated between one-another. So, instead of following Statistical approach, an Interlingua based approach for MT is more preferable in multi-lingual machine translation environment. UNL is an intermediate language, thus works on the concept of relations between words having attribute associated with it. By gaining meaning of token information, it is possible to interpret a sentence rather than translating it to another language. IAN framework is used to analyze a query. This system gives user a freedom to check the results in more than one language with a single interface and at the same time. This helps in shortens the language barriers and help in searching cross-lingual. We present our solution for performing cross lingual semantic based search. The project, that has been carried out in the thesis, present our semantic search engine of UNL, which encapsulate and sharpen previously proposed models. The use of UNL benefits us by providing feature of cross-lingual as well as multi-lingual search. We demonstrate that our model can predict search success more effectively than the existing state-of-the-art methods, on both our data and on a different set of log data collected from regular search engine sessions. For this different search strategies has been developed, which contributes in increasing recall of search results at the cost of decreasing precision, in manner of fall model. Together, our semantic search engine is another approach to perform not only a meaning based search but could also perform cross-lingual using IAN and EUGENE frameworks having dictionaries and rules of different languages.