Today: Clairvoyance Corporation enters a new chapter of technology development and licensing. The Company is focusing its efforts on research, prototyping, and intellectual property development, while concentration on extending its extensive suite of technology and charting new directions in unstructured text management.

Clairvoyance continues to develop and configure the most advanced technology for applications involving the processing of the language in documents. Our work addresses the problem of automatic organization and analysis of information, where speed, high accuracy, and flexibility are required. We are striving to make our technology widely available through licensing and partner relationships.

Clairvoyance Timeline

1988 - 1992: The CLARIT Project was begun in September 1988 in the Laboratory for Computational Linguistics at CMU under Dr. Evans' direction. The goals of the Project, sponsored principally by Digital Equipment Corporation, were to develop new approaches to large-scale information management, combining perspectives from natural-language processing (NLP), information science and artificial intelligence. The Project extended Dr. Evans' previous work in the areas of medical informatics, knowledge-based systems and computational linguistics.

By the end of 1991, the CLARIT Project had developed significant, new technology, including practical, robust automatic indexing using linguistic units; effective combination of complex index terms with advanced retrieval methods (such as vector-space modeling); automated thesaurus discovery (the ability to find the concepts that characterize a collection of documents); and a variety of related processes for analyzing texts. At this point, interest in CLARIT was high; several licenses for run-time versions of CLARIT system modules were sold by CMU to North American and European groups.

1992 - 1996: Based on the success of the Project, the obvious demand for commercial-grade systems for large-scale information management, and the proprietary innovations in the CLARIT system, Dr. Evans initiated the creation of CLARITECH. Mr. Archibold, a longtime friend of Dr. Evans, and a former researcher on the staff of the CLARIT Project, provided the principal seed capital for the venture. Additional contributions of resources, including license agreements, were made by CMU.

CLARITECH's original location

Shortly after CLARITECH's formation, several developers were recruited to the core staff, including key members of the CMU CLARIT Project team. During this period, Dr. Evans served the company as Chairman of the Board and Chief Scientist and led the company's product development and marketing efforts. Early adopters of CLARITECH software products included research institutes and government intelligence agencies.

In 1993, the National Technology Transfer Center (NTTC), a clearinghouse chartered to provide public access to information about federal research and technology, engaged CLARITECH to develop a comprehensive document and information delivery system. The finished system encompassed text scanning, records management and CLARIT NLP-based information retrieval over compound documents—page images coupled with OCR text representations. Information stored in distributed (and distant) databases was uniformly and rapidly accessed via Web browser and Windows interfaces. The Environmental Protection Agency (EPA) has subsequently adopted this system to support Internet-based search and delivery of page images of EPA reports.

In 1994, among other activities, CLARITECH undertook a project with the Carnegie Mellon University Libraries to develop the Heinz Electronic Library Interactive On-line System (HELIOS) to provide electronic access to some one million congressional papers of the late U.S. Senator John Heinz. The Senator's papers were scanned, converted to electronic text and indexed via the CLARIT retrieval application. The project resulted in the first content-searchable online archive with comprehensive free-text search.

Beginning also in 1994, CLARITECH began licensing its core technology in the form of a Toolkit—a suite of software modules designed to interoperate seamlessly to provide a wide variety of advanced text management and text analysis functions. The functions supported by the tools include:

• NLP (for Morphological Analysis and Phrase/Sub-Phrase Identification)
• Automatic Document Indexing
• Information Retrieval
• Routing/Filtering of Streaming Texts
• Subject Classification
• Summarization of Documents
• Automatic Thesaurus Discovery
• Spell-Checking (Empirically Based, without a Reference Lexicon)
• Information Extraction (Identification of Entities and Relations)
• "Virtual Hypertext"—Automatic Linking of Content “on the fly”
• Document & Concept Clustering
• Compound-Document Management (combining text and page images)

A distinguishing feature of CLARIT Tools is their use of (very fast!) NLP in virtually all text processing. The resulting deeper analysis of content and concepts leads to unprecedented precision and robustness.

Using such tools, CLARITECH and its customers in this period configured a variety of custom and replicable systems. These systems addressed many different kinds of problems, including large-scale newsfeed summarization and routing; information extraction and relation finding in a real-time analysis environment; and, of course, very advanced information retrieval.

In 1995, CLARITECH created a web-based application called CLARITWeb, an Internet system for accessing distributed information encompassing both text and images. Clients using CLARITWeb were able to perform their own specialized searches with CLARIT tools forming the core of an Internet service. An early user of CLARITWeb was Alexus International, a human resources management firm in Maryland. Alexus would scan thousands of resumes into databases at a restricted-access web site and their clients would use CLARITWeb technology to select specific individual documents from the entire group using sophisticated criteria of their own choosing. Alexus' customers thus received direct access to large numbers of candidates' resumes over the net, and all processing (and information enhancement) and analysis was provided as a service, giving Alexus' customers the benefit of very advanced technology without requiring the purchase, installation or maintenance of software.

1996 - 2000: In the spring of 1996, CLARITECH began considering opportunities for growth and new marketing partnerships. In the summer of 1996, CLARITECH and Justsystem Corporation of Tokushima, Japan, completed an agreement under which Justsystem Incorporated, the U.S.-based holding company of Justsystem Corporation, acquired a majority interest in CLARITECH. As a result, CLARITECH became a member of the family of Justsystem Group companies.

Clairvoyance's current location

Founded in 1979, Justsystem became the leading software developer in Japan in the late 1980s and early 1990s. Justsystem's core product, ATOK, solved the keyboard-character-to-Kanji conversion problem and is now the premier technology for Japanese language applications. Along with ATOK, Justsystem has developed a wide variety of productivity software and services, including the popular Ichitaro word-processing package, Hanako drawing program and other products for business accounting, databases and groupware, with sales exceeding 10 million units.

Justsystem and CLARITECH jointly developed several end-user applications and enterprise systems, including the Japanese-language search system, ConceptBase, which won the 1998 "Software Product of the Year" award from the Japanese Government. ConceptBase sales are expected to equal sales of Justsystem's desktop software within three years.

Other Justsystem/CLARITECH applications in the Japanese market include CB Summarizer and CB Classifier. In addition, core ConceptBase technology is incorporated in Japanese versions of Lotus Notes and MS Exchange. New products are being developed for enterprise and internet applications.

CLARIT Miner - Advanced Text Mining

* Note: CLARITECH Corporation was originally founded under the name "CLARIT Corporation". To avoid confusion with other existing trademarks and corporation names, the company name was changed to "CLARITECH Corporation" in 1994. The names "CLARIT" and "CLARITECH" are trademarks of CLARITECH Corporation.