Linguistic engineering approach to the enhancement of web-searching

This paper describes a proposal to improve the web document retrieval by facing the main three problems in document searching: first, traditional web search engines miss documents that are relevant to the user query and retrieve many that are not. Second, the query formulation is not as accessible as it could be, and some users have difficulties in expressing boolean queries. Third, the result sorting is usually arbitrary or based on statistical data. The on-going research presented in this paper focuses on the use of a natural language interface for interacting with search engines, both before and after the search process. This interface performs enrichment of user queries based on lexical resources, and makes use of semantic domain information for expanding and sorting the results. Preliminary experimental work and discussion of the obtained results are also outlined.