A Sentence Understander : Natural Language Query Interpreter
The "Sentence Understander”, or Natural Language Query Interpreter, will map an English source language into a target representation. The domain of discourse is limited to a form of database query or document retrieval. Data can be retrieved based on parameters such as; Author, SubJect, Date, Publication Type. For example, some of the sample requests the NLGI can process are: 'List the titles of books written by Aho'; 'Print everything about Natural Language'. An interesting feature of the Natural Language Query Interpreter is its ability to disambiguate automatically, the seeping of conjunctions in the natural language query in many cases. This is done by using semantic information provided by the data base. It should be pointed out that the main purpose of this project is to transform a natural language query into some "mechanical" query language, which can be interpreted by a separate query interpreter. Optimizations of queries or efficiency and structure of the underlying database will not be explicitly addressed.