April 15, 2015

Information Extraction and Dissemination

Web searching is great – if you know what you’re looking for and spend all your time searching. Cyc’s natural language understanding and knowledge modeling capabilities can help you do better by more completely understanding the content of text documents and proactively identifying items of interest to specific information consumers.

  • Extract critical information from text:  Cyc goes beyond simply tagging named entities (e.g., people, places, and companies); it can identify references to thousands of types of entities, including general concepts and those specific to a particular domain, as well as the relationships among them. Using an extensible set of semantic licensing rules, Cyc can use its understanding of the identified concepts and the document context to reduce ambiguity, improving the accuracy of the information extraction process. Furthermore, Cyc can create a formal model of the document content, enabling Cyc’s knowledge and reasoning to be applied to the information contained within the text.
  • Identify and disseminate documents based on user’s interests:  Using rich models of individuals, teams, or organizations, combined with a powerful and adaptable set of interest-matching rules, Cyc can identify what document content is likely to be of interest to a given recipient even if the information was not specifically requested. This is similar to, but more powerful than, search engine “alerts” because it can infer potential relevance based on knowledge about the recipient rather than relying solely on a set of user-specified search terms.