Method and apparatus for making predictions about entities represented in documents

• US 6,728,695 B1
• Filed: 05/26/2000
• Issued: 04/27/2004
• Est. Priority Date: 05/26/2000
• Status: Expired due to Term

• Current Assignee: Burning Glass International Incorporated
• Original Assignee: Burning Bush Technologies LLC
• Inventors: Crooks, Theodore J., Chow, Edmond D., Pathria, Anu K., Freitag, Dayne B., Laffoon, Mark A., Gopinathan, Krishna M.
• Primary Examiner(s): Breene, John
• Assistant Examiner(s): Wassum, Luke S
• Application Number:
  US09/589,180
• Time in Patent Office: 1,432 Days
• Field of Search: 706/21, 707/1-5, 707/10
• US Class Current: 1/1
• CPC Class Codes: G06F 16/313 Selection or weighting of terms for indexing
• CPC Class Codes: G06F 40/216 using statistical methods
• CPC Class Codes: G06F 40/289 Phrasal analysis, e.g. finite state techniques or chunking
• CPC Class Codes: G06F 40/35 Discourse or dialogue representation
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing
• CPC Class Codes: Y10S 707/99954 Version management

Abstract

A method and apparatus is disclosed for making predictions about entities represented in documents and for information analysis of text documents or the like, from a large number of such documents. Predictive models are executed responsive to variables derived from canonical documents to determine documents containing desired attributes or characteristics. The canonical documents are derived from standardized documents, which, in turn, are derived from original documents.

Claims

1. A method for creating a dataset for training predictive models from documents unraveled by considering objects in a time-ordered series and earlier objects in the series to be matched input/output pairs.

2. The method according to claim 1, wherein input to the output object includes parts of the document that are not in the time-ordered series.

3. The method according to claim 1, further including moving tags that occur in incorrect places with respect to the format of the document.

4. The method according to claim 1, further including checking fields to make sure the predicted contents have the right structure.

5. The method according to claim 1, further including comparing extraction of an HMM'"'"'s with those made by other methods.