Method and apparatus for producing an abstract of a document

• US 5,077,668 A
• Filed: 09/28/1989
• Issued: 12/31/1991
• Est. Priority Date: 09/30/1988
• Status: Expired due to Term

• Current Assignee: Toshiba Corporation
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Doi, Miwako
• Primary Examiner(s): Shaw, Dale M.
• Assistant Examiner(s): Brutman, Laura
• Application Number:
  US07/413,605
• Time in Patent Office: 824 Days
• Field of Search: 364/419
• US Class Current: 715/206
• CPC Class Codes: G06F 16/345 Summarisation for human users
• CPC Class Codes: G06F 40/174 Form filling; Merging

Abstract

A method and an apparatus for producing an abstract of a document capable of producing concise abstract with correct meaning precisely indicative of the content of the document automatically. The method includes the steps of: listing hint words which are preselected words indicative of presence of significant phrases that can reflect content of the document; searching all the hint words in the document; extracting sentences of the document in which any one of the listed hint words is found by the search; and producing an abstract for the document by juxtaposing the extracted sentences. An apparatus for performing this method is also disclosed.

Claims

1. A method of producing an abstract for a document, comprising the steps of:

listing hint words which are preselected words indicative of presence of significant phrases that can reflect content of the document;

searching all the hint words in the document;

extracting sentences of the document in which any one of the listed hint words is found by the search; and

producing an abstract for the document by juxtaposing the extracted sentences.

2. The method of claim 1, further comprising the step of listing structural attribute for each sentence of the document;

and wherein the search of some of the hint words are performed for only a fraction of all the sentences of the document, in accordance with nature of the hint words and the structural attributes of the sentences.

3. The method of claim 1, further comprising the steps of:

listing most preferable part of speech for each hint word; and

morphologically analyzing the sentences of the document; and

wherein the extraction of the sentences is limited to those sentences in which any one of the listed hint words that is found by the search is in the listed type of part of speech.

4. The method of claim 1, further comprising the step of modifying the extracted sentences such that for a hint word that is found with the listed type of part of speech which is a verb, a new sentence in which that hint word appears as a predicate will be created and replaces an original extracted sentence, while when a subject of the extracted sentence is a pronoun or missing, an appropriate sentence will be substituted for that pronoun, whereas for a hint word that is found with the listed type of part of speech which is a noun, a new sentence in which that hint word appears as a subject will be created and replaces the original extracted sentence.

5. The method of claim 1, further comprising the step of allowing an operator to edit the produced abstract.

6. An apparatus for producing an abstract for a document, comprising:

means for listing hint words which are preselected words indicative of presence of significant phrases that can reflect content of the document;

means for searching all the hint words in the document;

means for extracting sentences of the document in which any one of the listed hint words is found by the search; and

means for producing an abstract for the document by juxtaposing the extracted sentences.

7. The apparatus of claim 6, further comprising means for listing structural attribute for each sentence of the document;

and wherein the search of some of the hint words are performed for only a fraction of all the sentences of the document, in accordance with nature of the hint words and the structural attributes of the sentences.

8. The apparatus of claim 6, further comprising:

means for listing most preferable part of speech for each hint word; and

means for morphologically analyzing the sentences of the document;

and wherein the extraction of the sentences is limited to those sentences in which any one of the hint words listed that is found by the search is in the listed type of part of speech.

9. The apparatus of claim 6, further comprising means for modifying the extracted sentences such that for a hint word that is found with the listed type of part of speech which is a verb, a new sentence in which that hint word appears as a predicate will be created and replaces an original extracted sentence, while when a subject of the extracted sentence is a pronoun or missing, an appropriate sentence will be substituted for that pronoun, whereas for a hint word that is found with the listed type of part of speech which is a noun, a new sentence in which that hint word appears as a subject will be created and replaces the original extracted sentence.

10. The apparatus of claim 6, further comprising means for allowing an operator to edit the produced abstract.