Concept identification system and method for use in reducing and/or representing text content of an electronic document

• US 6,823,331 B1
• Filed: 08/28/2000
• Issued: 11/23/2004
• Est. Priority Date: 08/28/2000
• Status: Expired due to Fees

• Current Assignee: Entrust Incorporated (Entrust Corp.)
• Original Assignee: Entrust Incorporated (Entrust Corp.)
• Inventors: Abu-Hakima, Suhayya
• Primary Examiner(s): Wassum, Luke S
• Application Number:
  US09/649,028
• Time in Patent Office: 1,548 Days
• Field of Search: 707/3-6
• US Class Current: 707/758
• CPC Class Codes: G06F 16/34 Browsing; Visualisation therefor
• CPC Class Codes: G06F 16/353 into predefined classes
• CPC Class Codes: Y10S 707/956 Hierarchical
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching

Abstract

A concept identification system useful in reducing and/or representing text content of an electronic document and in highlighting the content of the document. A concept knowledge base comprises a plurality of concepts and each concept comprises one or more subconcepts linked to each other and to the concept on a hierarchical basis. One or more of the subconcepts may be linked to one or more subconcepts of another concept. A concept matching module matches text of the document to subconcepts of the concept knowledge base and assesses any links between the matched subconcepts and other concepts and/or subconcepts of the concept knowledge base. From this a determination is made of whether the document relates to a concept of the knowledge base. With an identification of such concept a document representation generator may produce a precis of the document based on a template associated with such concept. For highlighting of a document a highlighter module determines key content of the input document and an interface integrates the concept identification system and the highlighter module. An output module produces an output highlight document from the key content.

Claims

1. A computer-readable concept identification system including modules executable by said computer'"'"'s programmable processor for identifying a concept to which an electronic document relates, said concept identification system comprising:

(a) a concept knowledge base comprising a plurality of concept schemas wherein each said concept schema comprises;

(i) a concept comprising concept terms, including synonyms, that represent said concept; and

(ii) a plurality of subconcepts linked to said concept and/or to each other, on a hierarchical basis, and comprising subconcept terms, including synonyms, that represent said subconcept; and

wherein said concept schemas comprise one or more sets of multi-relationship concepts wherein one or more subconcepts of a concept of said multi-relationship concepts of each said set is linked to another concept of said multi-relationship concepts of said set through said hierarchically linked subconcepts and concepts of said multi-relationship concepts of said set; and

,

(b) a concept matching module configured for;

(i) comparing key word(s) and/or key phrase(s) and/or key sentence fragment(s) of said document to said concept terms and subconcept terms of said concept schemas and identifying matched terms from said comparing;

(ii) counting said matched terms to determine a match count;

(iii) identifying matched multi-relationship concepts from any subconcepts of multi-relationship concepts comprising said matched terms;

(iv) firstly assigning threshold weights to only those of said matched terms which are not comprised in said matched multi-relationship concepts, wherein said firstly assigned threshold weight assigned to each said matched term is based on a level of inherent distinctiveness of said matched term to said concept of said concept schema comprising said matched term;

(v) determining which of said multi-relationship concepts is more related to said document on the basis of said firstly assigned threshold weights;

(vi) secondly assigning threshold weights to said matched key word(s) and/or key phrase(s) and/or key sentence fragment(s) which are matched to terms of subconcepts of said multi-relationship concepts on the basis of said multi-relationship concept determined to be more related to said document;

(vii) for each said concept schema having said matched terms, calculating an overall matching weight representative of said match count and said assigned threshold weights; and

,

(viii) comparing each said overall matching weight calculated for a concept schema to a predetermined matching weight for that concept schema, and from said comparing, determining whether said document is characterized by the concept of said that concept schema.

2. A computer-readable document interpretation system including modules executable by said computer'"'"'s programmable processor for highlighting the content of an electronic input document and producing therefrom an electronic output highlight document, said system comprising:

(a) a concept identification system according to claim 1 for providing said identified concept of said concept knowledge base which characterizes said input document;

(b) a highlighter module interfaced with said concept identification system and configured for determining key content of said input document, said highlighter module comprising a comparing module for comparing content of said input document to said concept schema for said identified concept and for determining said key content on the basis of said concept terms and subconcept terms including a hierarchical position of said terms, in said concept schema; and

,

(c) an output module configured for producing said output highlight document from said key content.

3. A system according to claim 2 and further comprising a document mapping module configured for producing a static document map of said input document'"'"'s content, wherein said highlighter module applies to said static document map weightings derived from determinations made by said comparing module.

4. A concept identification system according to claim 1 wherein said secondly assigned threshold weights are predetermined on the basis of whether said concept is normally described by distinctive or non-distinctive words/phrases/sentence fragments.

5. A system according to claim 4 and further comprising a document representation generator for producing a precis of said document based on a template associated with said concept identified to characterize said document.

6. A system according to claim 5 and further comprising an output module for communicating an identification of said concept identified by said matching module to characterize said document.

7. A method for identifying a concept to which an electronic document relates of an electronic document for use in reducing and/or representing text content of said document, said method comprising:

(a) providing a concept knowledge base comprising a plurality of concept schemas wherein each said concept schema comprises;

(i) a concept comprising concept terms. Including synonyms, that represent said concept; and

(ii) a plurality of subconcepts linked to said concept and/or to each other, on a hierarchical basis, and comprising subconcept terms, including synonyms, that represent said subconcept; and

wherein said concept schemas comprise one or more sets of multi-relationship concepts wherein one or more subconcepts of a concept of said multi-relationship concepts of each said set is linked to another concept of said multi-relationship concepts of said set through said hierarchically linked subconcepts and concepts of said mufti-relationship concepts of said set; and

,

(b) comparing key word(s) and/or key phrase(s) and/or key sentence fragment(s) of said document to said concept terms and subconcept terms of said concept schemas and identifying matched terms from said comparing;

(c) counting said matched terms to determine a match count;

(d) identifying matched multi-relationship concepts from any subconcepts of multi-relationship concepts comprising said matched terms;

(e) firstly assigning threshold weights to only those of said matched terms which are not comprised in said matched multi-relationship concepts, wherein said firstly assigned threshold weight assigned to each said matched term is based on a level of inherent distinctiveness of said matched term to said concept of said concept schema comprising said matched term;

(f) determining which of said multi-relationship concepts is more related to said document on the basis of said firstly assigned threshold weights;

(g) secondly assigning threshold weights to said matched key word(s) and/or key phrase(s) and/or key sentence fragment(s) which are matched to terms of subconcepts of said multi-relationship concepts on the basis of said multi-relationship concept determined to be more related to said document;

(h) for each said concept schema having said matched terms, calculating an overall matching weight representative of said match count and said assigned threshold weights; and

,

(i) comparing each said overall matching weight calculated for a concept schema to a predetermined matching weight for that concept schema, and from said comparing, determining whether said document is characterized by the concept of that concept schema.

8. A method for highlighting the content of an electronic input document and producing therefrom an electronic output highlight document, said method comprising:

(a) identifying a concept of said concept knowledge base which characterizes said input document according to claim 7;

(b) determining key content of said input document including comparing content of said input document to said concept schema for said identified concept and determining said key content on the basis of said concept terms and subconcept terms, including the hierarchical position of said terms, in said concept schema; and

,

(c) producing said output highlight document from said key content.

9. A method according to claim 8 and further comprising producing a static document map of said input document'"'"'s content, whereby said step of determining key content includes applying to said static document map weightings derived from said step of comparing content.

10. A method according to claim 7 whereby said secondly assigned threshold weights are predetermined on the basis of whether said concept is normally described by distinctive or non-distinctive words/phrases/sentence fragments.

11. A method according to claim 10 and further comprising producing a precis of said document based on a template associated with said concept identified to characterize said document.

12. A method according to claim 11 and further comprising communicating said concept identified to characterize said document.