Language Systems Inc. 5959 Topanga Canyon Blvd. Suite 340 Woodland Hills, CA 91367 Phone: (818)703-5034 Email: chris@lsi.com
Dr. Montgomery, who is President of LSI and founder of the corporation, has more than 20 years of research experience in linguistic and natural language processing systems, with a special focus on the artificial intelligence area of language understanding by computers. She received her B.S. (French) and M.S. (Linguistics) degrees from Georgetown University, and her Ph.D. from UCLA (Anthropological Linguistics). Recent activities include natural language and speech understanding, cognitive modeling of users, and intelligent database systems. Currently, she is directing projects in speech and natural language understanding, including a voice-to-voice translation project which converts spoken English queries into Spanish, as well as translating the Spanish responses into spoken English.
Dr. Montgomery is a member of the ACM Publications Board. She is a former editor of the Artificial Intelligence and Language Processing Department of the Communications of the ACM, a former member of the ACM SIG Board, and is past Chairman of SIGLASH (The ACM Special Interest Group for Language Analysis and Studies in Humanities).
When the concept of translating from one language to another with a computer first appeared in the 1950s, the notion that a machine might be able to perform the uniquely human functions of understanding natural language input and generating natural language output -- in a SECOND language yet!! -- served primarily as a source of material for stand-up comedians. However, natural language processing technology has advanced considerably since the initial days of uniformly bad machine translation. The presentation analyzes progress since that time and describes the components of current natural language processing systems which have the objective of full text understanding.
As natural language processing technology assumes a more important role in user interfaces and data/knowledge acquisition for automated information processing systems, questions arise as to the accuracy and reliability of this technology.
The presentation provides an overview of natural language processing system components and describes "black box" and "glass box" types of evaluation that have been applied to natural language processing systems, concluding with an interpretation evaluation results.
On the frontier of current information processing technology is the integration of speech and natural language processing. Applications for these spoken language systems include voice interfaces for a variety of information processing systems, ranging from on-line data bases to diagnostic experts, as well as applications such as machine translation, intelligent tutoring systems, dictation and annotation of texts. The presentation discusses current technology in speech and natural language processing, and the integration of these technologies into systems for understanding and generation of spoken language.