Share Site Map Contact us Home Page
Home > Background > AI at the turn of the millennium > Corporate AI

History of AI

AI at the turn of the millennium
Academic AI
Corporate AI
Commercial AI

Speaking Machines

From Myth to Reality part 1

From Myth to Reality part 2

From Myth to Reality part 3

From Myth to Reality part 4

From Myth to Reality Full
 Corporate AI
Microsoft Research - Natural Language Processing Group
IBM Research
The CYC Project
Microsoft Research - Natural Language Processing Group
The Natural Language Processing (NLP) group at Microsoft is working to design and build a computer system that analyzes and understands regular human language. The group is working with 7 languages: Chinese, English, French, German, Japanese, Korean, and Spanish.

Some results of the group's research: the English grammar checker in MS Office, the Encarta Encyclopedia product's question-answering utility, and natural language queries in SQL 7.

The Machine Learning and Applied Statistics (MLAS) Group: The MLAS group is building software that automatically learns from data. The software will help software perform intelligent tasks like handwriting recognition and natural language processing, and help human data analysts make better sense of their data.

The Natural Language Computing Group: The Natural Language Computing group is exploring advanced natural language technologies, particularly for the Chinese language. The Chinese Corpus Collection, Classification and Annotation project, for example, is an effort to build a large Chinese text corpus as the infrastructure of statistical language modeling, in which conclusions about a text are drawn by automatically analyzing large bodies of text.

The Speech Technology Group: The Speech Technology group researches and develops spoken language technologies. The major area of research is the Dr. Who project. The goal of Dr. Who is to build applications that make computers available everywhere, especially in the Next Generation Windows Services (NGWS). The first application of Dr. Who is the MiPad project, a personal information manager used with speech and pen.

Mindnet: Developed by the NLP (Natural Language Processing) group at Microsoft Research, Mindnet is a system for automatically extracting a massively hyperlinked web of concepts. Mindnet tries to tie together all of human understanding in words; it obtains information by automatically sorting and indexing texts. Microsoft has successfully extended the MindNet methodology to the processing of other text databases, such as the Encarta 98 Encyclopedia.
IBM Research
Several projects within IBM Research deal with computational linguistics and technologies such as speech processing, machine translation, universal and application-specific dialog engines, information retrieval, automatic text summarization, natural language understanding and generation.

In particular, the Human Language Technology Group develops conversational interfaces based on speech recognition and NLP techniques (e.g. parsing).

IBM has effectively turned its natural language research into saleable products, sold via IBM's Speech Business Unit and Industry Solution Unit, Edmark, Lotus, as well as independent software vendors, OEMs, and value added resellers.
The CYC Project
Cycorp, based in Austin, Texas, is the world’s largest common sense project and has been under development since 1984. CYC is creating a huge knowledge base of interlinked information, comprised of over 1,000,000 hand-coded "rules" or assertions. Given enough assertions, CYC believes that a computer will be capable of human-level reasoning.

CYC currently produces CycAnswers, a knowledge-based question-answering system, and CycSecure, a non-invasive network scanning system that deduces the steps a hacker must take to attack a company's network. Cycorp is planning to make some of its technology available for public use in August 2001, via an open source initiative.

In 1995, CYC spun off from the Microelectronics and Computer Technology Corporation (MCC), U.S.A. Recently, CYC became involved with the World Wide Web Council's "Semantic Web" initiative, announcing that it will open its "ontology" to the larger web community.

Doug Lenat is the founder and president of Cycorp.

The Cyc Open Source Project is at

  Printable version