System, method and computer program product for performing unstructured information management and automatic text analysis, and providing multiple document views derived from different document tokenizations

• US 7,139,752 B2
• Filed: 05/30/2003
• Issued: 11/21/2006
• Est. Priority Date: 05/30/2003
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Zadrozny, Wlodek W, Ciccolo, Arthur C, Carmel, David, Mass, Yosi, Soffer, Aya, Maarek, Yoelle, Broder, Andrei Z, Ferrucci, David
• Primary Examiner(s): Robinson, Greta
• Application Number:
  US10/449,409
• Publication Number:
  US 20040243645A1
• Time in Patent Office: 1,271 Days
• Field of Search: 707/5, 707/10, 707/201, 707/1, 707/3, 707/4, 707/100, 704/9, 704/7, 715/500, 715/542
• US Class Current: 1/1
• CPC Class Codes: G06F 16/313 Selection or weighting of terms for indexing
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

Disclosed is a system architecture, components and a searching technique for an Unstructured Information Management System (UIMS). The UIMS may be provided as middleware for the effective management and interchange of unstructured information over a wide array of information sources. The architecture generally includes a search engine, data storage, analysis engines containing pipelined document annotators and various adapters. The searching technique makes use of a two-level searching technique. Also disclosed is system, method and computer program product to process document data. The method includes inputting a document and operating at least one text analysis engine that comprises a plurality of coupled annotators for tokenizing document data for identifying and annotating a particular type of semantic content. Operating the at least one text analysis engine generates a plurality of views of a document, where each of the plurality of views are derived from a different tokenization of the document. The method further includes storing the plurality of views in a common data structure associated with the document.

Claims

1. A data processing system for processing document data, comprising:

data storage for storing a collection of document data that comprises unstructured document data; and

at least one text analysis engine that comprises a plurality of coupled annotators at least some of which are operable for tokenizing document data for identifying and annotating a particular type of semantic content;

where said at least one text analysis engine operates to generate a plurality of views of a document, each of said plurality of views being derived from a different tokenization of the document.

2. A data processing system as in claim 1, where said data processing system operates with an inverted file system for storing, for each of said views, said annotations, a list comprising occurrences of respective annotations and, for each listed occurrence of a respective annotation, a set comprised of a plurality of token locations spanned by at least one of said annotations.

3. A data processing system as in claim 1, further comprising a search engine that is responsive to a query that comprises a logical combination of at least two predicates, where a first predicate pertains to a first view and a second predicate pertains to a second view, said search engine returning from said data storage at least one document that satisfies the logical combination of predicates.

4. A data processing system as in claim 1, where there are a plurality of said text analysis engines, further comprising an analysis structure broker coupled to said plurality of text analysis engines for coordinating communication between individual ones of the text analysis engines and a common abstract data structure.

5. A data processing system as in claim 4, where said common data structure comprises an object-based representation implemented as a hierarchical tree-like structure that supports single inheritance.

6. A data processing system as in claim 4, where said common data structure comprises an object-based representation implemented as a directed graph structure that supports multiple inheritance.

7. A computer-implemented method to process document data, comprising:

inputting a document; and

operating at least one text analysis engine that comprises a plurality of coupled annotators for tokenizing document data for identifying and annotating a particular type of semantic content;

where operating said at least one text analysis engine generates a plurality of views of a document, each of said plurality of views being derived from a different tokenization of the document;

further comprising

storing said plurality of views in a common data structure associated with the document.

8. A method as in claim 7, where said common data structure implements an inverted file system for storing, for each of said views, said annotations, a list comprising occurrences of respective annotations and, for each listed occurrence of a respective annotation, a set comprised of a plurality of token locations spanned by at least one annotation.

9. A method as in claim 7, where a semantic search is responsive to a query that comprises a logical combination of at least two predicates, where a first predicate pertains to a first view and a second predicate pertains to a second view, for returning at least one document that satisfies the logical combination of predicates.

10. A method as in claim 7, where there are a plurality of the text analysis engines, further comprising coordinating communication between individual ones of the text analysis engines and said common abstract data structure.

11. A method as in claim 10, where said common data structure comprises an object-based representation implemented as a hierarchical tree-like structure that supports single inheritance.

12. A method as in claim 10, where said common data structure comprises an object-based representation implemented as a directed graph structure that supports multiple inheritance.

13. A computer program product embodied on a computer-readable medium and comprising program code for directing at least one computer to process document data, comprising:

a program code segment for inputting a document; and

a program code segment for implementing at least one text analysis engine that comprises a plurality of coupled annotators for tokenizing document data for identifying and annotating a particular type of semantic content;

where operation of said at least one text analysis engine generates a plurality of views of a document, each of said plurality of views being derived from a different tokenization of the document;

further comprising

a program code segment for storing said plurality of views in a common abstract data structure associated with the document.

14. A computer program product as in claim 13, where said common data structure implements an inverted file system for storing, for each of said views, said annotations, a list comprising occurrences of respective annotations and, for each listed occurrence of a respective annotation, a set comprised of a plurality of token locations spanned by at least one annotation.

15. A computer program product as in claim 13, where a semantic search is responsive to a query that comprises a logical combination of at least two predicates, where a first predicate pertains to a first view and a second predicate pertains to a second view, for returning at least one document that satisfies the logical combination of predicates.

16. A computer program product as in claim 13, where there are a plurality of the text analysis engines, further comprising a computer program code segment for coordinating communication between individual ones of the text analysis engines and said common abstract data structure.

17. A computer program product as in claim 16, where said common data structure comprises an object-based representation implemented as a hierarchical tree-like structure that supports single inheritance.

18. A computer program product as in claim 16, where said common abstract data structure comprises an object-based representation implemented as a directed graph structure that supports multiple inheritance.

19. A computer program product as in claim 13, where the document tokenization is derived from a plain text document.

20. A computer program product as in claim 13, where the document tokenization is derived from a language translation of the document.

21. A computer program product as in claim 13, where the document tokenization is derived from a summary of the document.

22. A computer program product as in claim 13, where the document tokenization is derived from a plain text variant of a marked-up document.

23. A computer program product as in claim 13, where the document tokenization is derived from a plain text variant of a HTML document.

24. A computer program product as in claim 13, where the document tokenization comprises an n-gram tokenization of the document.

25. A computer program product as in claim 13, where the document tokenization is derived from a multi-media document.

26. A computer program product as in claim 13, where the document tokenization is comprised of multi-media objects.

27. A computer program product as in claim 13, where the document tokenization is comprised of objects having different data types.