Method for recognizing compound terms in a document

• US 5,842,217 A
• Filed: 12/30/1996
• Issued: 11/24/1998
• Est. Priority Date: 12/30/1996
• Status: Expired due to Fees

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Light, John
• Primary Examiner(s): KULIK, PAUL V
• Application Number:
  US08/773,194
• Time in Patent Office: 694 Days
• Field of Search: 707/5, 707/101, 707/500, 707/530, 707/531
• US Class Current: 1/1
• CPC Class Codes: G06F 16/90344 by using string matching techniques
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation

Abstract

A method is provided for identifying compound terms in a document that is represented by a stream of tokens. The stream of document tokens is scanned for an initial term associated with a compound term and a compound term template is accessed when the initial term is identified. The template includes content, retention, and token specifications for the compound term. The stream of tokens is compared with the template, and when the stream matches the content specification of the template, a token representing the compound term is tagged according to the retention specification and added to the stream of tokens. The tagged token is stopped according to the retention specification represented by its tag.

Claims

1. A method for identifying a compound term in a document that is represented as a stream of document terms, the method comprising the steps of:

scanning the stream of document terms for an initial term associated with the compound term;

when the initial term is identified, accessing a compound term template that includes content, retention, and token specifications for the compound term;

comparing the stream, beginning with the initial term, to the compound term template; and

adding to the document stream a tagged token indicated by the token and retention specifications of the template, when the stream matches the content specification of the template.

2. The method of claim 1, wherein the accessing step comprises the substeps of:

converting the initial term into a location key; and

accessing a compound term template having an entry point specified by the location key.

3. The method of claim 1, wherein the comparing step comprises the substeps of:

identifying the content specification of the template; and

comparing the stream with the content specification.

4. The method of claim 3, wherein the identifying step comprises determining component terms, including a last term, of the compound term from the content specification and wherein the comparing step comprises comparing the document terms of the stream with the component terms until the last term is identified.

5. The method of claim 1, wherein the step of accessing a compound term template comprises the substeps of;

accessing a compound term template;

determining from the content specification whether there is an alternative form template; and

accessing the alternative form template, including content, retention, and token specifications, when indicated by the content specification of the compound term template.

6. The method of claim 5, wherein the comparing step comprises comparing the stream, beginning at the initial term, with the content specifications of the compound and alternative form templates.

7. A method for identifying a compound term in a document, using a data structure that represents the compound term, the method comprising the steps of:

converting the document into a stream of document terms;

scanning the stream of document terms for an initial term associated with the compound term;

when the initial term is identified, comparing the stream, beginning with the initial term, to content indications specified by the data structure; and

when the content indications are matched by the stream, tagging a compound term token in accordance with status indications specified in the data structure and adding the tagged compound term token to the stream.

8. The method of claim 7, wherein the comparing step comprises the substeps of:

determining from the data structure, whether the compound term includes alternative forms; and

when the data structure indicates that alternative forms exist, determining alternative content indications in addition to those specified by the data structure and comparing the stream, beginning with the initial term to the content indications and the alternative content indications.

9. The method of claim 7, including the additional step of stopping the compound term token in accordance with the tag.

10. The method of claim 9, wherein the stopping step comprises the substeps of

eliminating component terms from the stream and retaining the compound term token when the tag has a first value;

and

retaining the component terms and the compound term token in the stream when the tag has a second value.

11. A system for identifying compound terms in a document that is represented as a stream of document terms, the system comprising:

a comparison engine coupled to receive the stream of document terms, for comparing the received document terms to a data structure representing the compound term; and

a data structure representing the compound term and coupled to the comparison engine through a location derived from the initial term, the data structure including a content specification for indicating a component term of the compound term, a retention tag associated with the content specification for indicating a status of the component term in an index representation of the document; and

a token specification, associated with the content specification for identifying a token to be added to the document stream the content specification is met.

12. A method for identifying a compound term in a document, using a data structure that represents the compound term, the method comprising the steps of:

tokenizing the document into a stream of document terms;

detecting an initial term of the compound term in the stream of document terms;

identifying content and retention specifications for the compound term from the data structure;

adding a token representing the compound term to the stream when the specified content indication is matched by the stream; and

tagging the token according to the retention specification indicated in the data structure.