This Small Business Innovation Research (SBIR) Phase I project addresses the construction of a cross-lingual retrieval system based on deep linguistic analysis intended to deliver efficiencies and cost savings and reduce communication barriers worldwide. By 2003, 70 percent of the estimated 1.3 billion users of the Internet will be non-English speaking. Thus the need to enable automated, reliable accessibility to information in more than one language is clear. Most existing information retrieval systems are either monolingual or cross-language with limited success and no translation capabilities. The investigator's Cross-Language Information Retrieval System (CLIR) will produce significant performance increases in the recall and precision of a cross-language sear. The investigative team will encapsulate cutting edge natural language processing (NLP) technology developed on a consistent theoretical foundation into a modular and scalable application-programming interface. The system will incorporate an extraction component for the matching of relevant lexemes based on their syntactic context, and a probabilistic module, which will disambiguate multiple interpretation based on expectation. The firm has secured an exclusive license from Russian Academy of Sciences for the use of complex NLP algorithms that will be integrated into the CLIR technology, creating a state-of-the-art linguistic kernel.Universal Dialog proffered technology will provide any person, business, organization or government, as well as any software that integrates their Cross-Language Information Retrieval, a faster, less costly solution for the retrieval of multilingual information, with a higher degree of precision, recall and scalability that CLIR methods currently available. The firm plans to integrate the technology to be developed in this project into its Cross-Language Communication Platform (CLCP). CLCP will enable users to retrieve, summarize, and fully translate retrieved data.
