This Small Business Innovation Research (SBIR) Phase I project submitted by Teragram Corporation will design and prototype critical theoretical and algorithmic elements of a linguistic indexing and search software using finite-state technology. Prototyping the architecture of a search engine will enable searching and indexing syntactically and semantically related concepts. This research project embodies linguistic and finite-state techniques for the purpose of search and indexing of textual data. This prototype could lead to a new generation of search and indexing technologies, which will enable better search by accessing linguistically related concepts, while being flexible, fast and scalable to huge amounts of textual information. The company's linguistic technology is based on formal and algorithmic properties of finite-state automata and finite-state transducers and enables linguistic processing at very high speeds while achieving very high data compression rates. Finite-state technologies have tremendous potentials in linguistic processing for indexing purposes. If the proposed research meets its goals, the resulting software will have numerous commercial applications such as file system indexing, document indexing, and document retrieval and Internet search engines.
