×

Method and system for information extraction and modeling

  • US 7,890,533 B2
  • Filed: 05/17/2006
  • Issued: 02/15/2011
  • Est. Priority Date: 05/17/2006
  • Status: Active Grant
First Claim
Patent Images

1. A method for visually modeling information sought from a set of documents implemented using a computer having a processor and a display, comprising:

  • identifying a set of documents;

    applying a filter to the set of documents to produce raw text;

    analyzing the raw text using a lexica module and a POS (part of speech) tagger by operation of the processor;

    creating a set of POS (part of speech) tagged documents based on the analysis of the raw text, the set of POS (part of speech) tagged documents corresponding to the set of documents;

    presenting the analysis of the raw text to a user;

    creating a plurality of concepts based on the analysis of the raw text;

    creating a visual model comprising visual elements corresponding to the plurality of concepts;

    presenting the visual model to the user on the display;

    enabling the user to add a new visual element to the visual model, the new visual element corresponding to a new concept;

    enabling the user to add a new relation between visual elements in the visual model, the new relation between visual elements representing a new relation between concepts corresponding to the visual elements;

    receiving a definition of a concept from the user via a selection of a visual model corresponding to the concept;

    generating extractors, each extractor corresponding to one of the visual elements or the relations between the visual elements in the visual model;

    based on a user selection of one of the visual elements or the relations, extracting a POS (part of speech) tagged document from the set of POS (part of speech) tagged documents using the corresponding extractor, the extracted POS (part of speech) tagged document containing information related to the concept corresponding to the selected visual element or the selected relation;

    presenting the extracted POS (part of speech) tagged document to the user;

    customizing the visual model based on user input in response to the extracted POS (part of speech) tagged document; and

    exporting the customized model.

View all claims
  • 2 Assignments
Timeline View
Assignment View
    ×
    ×