Cormine Intelligent Data

SBIR DKB:

The Small Business Innovation Research (or SBIR) program is a United States Government program designed to assist small businesses, provide competitive opportunities and stimulate innovation. As the program has grown and because of the wide-ranging nature of the research projects funded, it has become increasingly difficult for interested parties to find SBIR awards in a given topic area. This problem is made more urgent by a recent push by various agencies within the U.S. government to make more effective use of the SBIR program research. A number of agencies have developed conventional SBIR databases, but these efforts have each been hampered by the difficulty in organizing and making avaialble for search and retrieval a large number of highly technical publications across a wide variety of topics. CorMine has undertaken a number of projects using the SBIR program data.

Using a set of tailored spiders we captured approximately 40,000 SBIR reward documents and we’ve classified them using a variety of approaches for several projects such as:

  • Using the 40,000 SBIR projects we captured and an existing engineering taxonomy we developed a Dynamic Knowledge Base for use as an efficient search tool for engineers interested in finding SBIR programs that address particular topics.
  • Using the classifier tools to substantially enrich an existing engineering taxonomy, thus improving the taxonomy itself and making it more useful for internal and external applications.
  • Using various sets of densely technical, topic-focused documents as training sets, our researchers and analysts have uncovered a broad range of SBIR awards that relate to the training sets in known ways. Thus moving beyond the realm of producing marching documents to the realm of assembling document on and receiving answers about complex topics.

Once one of our Dynamic Knowledge Bases is created based on a given set of documents and well-ordered vocabulary, it continually and dynamically updates the classification system as changes are made and new documents added. Unlike traditional databases, our system won't lose its relevance, go out of date, or need continued human updating and maintenance
Frank Byrum
Chief Scientist
CorMine Intelligent Data

© Copyright 2006 by CorMine Intelligent Data, all rights reserved