Relational text index creation and searching

• US 6,741,988 B1
• Filed: 08/10/2001
• Issued: 05/25/2004
• Est. Priority Date: 08/11/2000
• Status: Expired due to Fees

• Current Assignee: Attensity Corporation (Attensity Group Incorporated)
• Original Assignee: Attensity Corporation (Attensity Group Incorporated)
• Inventors: Bean, David L., Wakefield, Todd D., Lorenzen, Jeffrey A.
• Primary Examiner(s): Mizrahi, Diane D.
• Assistant Examiner(s): Wu, Yicun
• Application Number:
  US09/928,249
• Time in Patent Office: 1,019 Days
• Field of Search: 707/1, 707/10, 707/101, 704/1, 704/4, 704/8, 704/9, 704/10, 345/708, 345/853, 700/104, 700/90, 341/51, 706/11, 703/6, 703/2
• US Class Current: 707/741
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: G06F 16/30 of unstructured textual data document management systems G06F16/93
• CPC Class Codes: G06F 16/313 Selection or weighting of terms for indexing
• CPC Class Codes: Y10S 707/99934 Query formulation, input preparation, or translation
• CPC Class Codes: Y10S 707/99937 Sorting

Abstract

In an environment where it is desire to perform information extraction over a large quantity of textual data, methods, tools and structures are provided for building a relational text index from the textual data and performing searches using the relational text index.

Claims

1. A computer program product located to one or more storage media devices usable to perform thematic role data mining on a set of documents, said computer program product comprising computer readable instructions executable by a computer to perform the functions of:

reading a relational text index, said index including thematic role information and corresponding document location information, the thematic role information being a product of thematic role extraction, the document location information including references to source documents containing the natural language text sourced for the thematic role extraction;

performing data mining analytic processing on thematic role information read from the relational text index in said reading, said processing identifying common events or attributes in the thematic role information of the relational text index as a product; and

providing the common event or attribute product for further processing or display to a user,

wherein said computer readable instructions are further executable to utilize sentence information to build a relational text index readable by said reading, the sentence information including thematic role information;

wherein said computer readable instructions are further executable to perform thematic role assignment on caseframe extractions to generate thematic role extraction information suitable for inclusion into a relational text index; and

wherein said computer readable instructions are further executable to perform unification of the thematic role extraction information.

2. The computer program product of claim 1, wherein said computer readable instructions are executable to read a relational text index containing thematic role information, references of source documents, and references to locations within source documents corresponding to the sources of thematic role extractions.

3. The computer program product of claim 1, wherein said computer readable instructions are further executable to utilize sentence information to build a relational text index readable by said reading, the sentence information including thematic role information.

4. The computer program product of claim 3, wherein said computer readable instructions are further executable to perform thematic role assignment on caseframe extractions to generate thematic role extraction information suitable for inclusion into a relational text index.

5. The computer program product of claim 4, wherein said computer readable instructions are further executable to perform subject-specific conceptual role assignment.

6. The computer program product of claim 4, wherein said computer readable instructions are further executable to:

parse natural language sentences contained in a set of source documents; and

apply caseframes to the parsed natural language sentences to generate caseframe extractions.

7. The computer program product of claim 6, wherein said computer readable instructions are further executable to convert a set of sourced documents to a set of formatted documents, the formatted documents having a format suitable for said parsing.

8. The computer program product of claim 6, wherein said computer readable instructions are further executable to collect documents from a source and present the collected documents for said converting.

9. The computer program product of claim 1, wherein said computer readable instructions are executable to read a relational text index containing thematic role information, the location of documents, and locations within the documents corresponding to thematic role information.

10. A computer program product located to one or more storage media devices usable to perform thematic role data mining on a set of documents, said computer program product comprising computer readable instructions executable by a computer to perform the functions of:

reading a relational text index, said index including thematic role information and corresponding document location information, the thematic role information being a product of thematic role extraction, the document location information including references to source documents containing the natural language text sourced for the thematic role extraction;

performing data mining analytic processing on thematic role information read from the relational text index in said reading, said processing identifying common events or attributes in the thematic role information of the relational text index as a product; and

providing the common event or attribute product for further processing or display to a user,

wherein said computer readable instructions are further executable to utilize sentence information to build a relational text index readable by said reading, the sentence information including thematic role information;

wherein said computer readable instructions are further executable to perform thematic role assignment on caseframe extractions to generate thematic role extraction information suitable for inclusion into a relational text index; and

wherein said computer readable instructions are further executable to perform subject-specific conceptual role assignment.

11. A computer program product located to one or more storage media devices usable to perform thematic role data mining on a set of documents, said computer program product comprising computer readable instructions executable by a computer to perform the functions of:

reading a relational text index, said index including thematic role information and corresponding document location information, the thematic role information being a product of thematic role extraction, the document location information including references to source documents containing the natural language text sourced for the thematic role extraction;

performing data mining analytic processing on thematic role information read from the relational text index in said reading, said processing identifying common events or attributes in the thematic role information of the relational text index as a product; and

providing the common event or attribute product for further processing or display to a user,

wherein said computer readable instructions are further executable to utilize sentence information to build a relational text index readable by said reading, the sentence information including thematic role information;

wherein said computer readable instructions are further executable to perform thematic role assignment on caseframe extractions to generate thematic role extraction information suitable for inclusion into a relational text index; and

wherein said computer readable instructions are further executable to;

parse natural language sentences contained in a set of source documents; and

apply caseframes to the parsed natural language sentences to generate caseframe extractions.