← Back to Patent Explorer
US5745602A Expired / Lapsed US ⏱️ 0.0 years remaining

Automatic method of selecting multi-word key phrases from a document

Assignee: Xerox Corp
•
Filing: 1995-05-01
•
Granted: 1998-04-28
•
Expires: 2015-05-01
📄 Original PDF 🌐 Google Patents 🔄 Sync Now

Abstract

Abstract An automatic method of generating key phrases for a machine readable document. The method begins by breaking the text of the document into multi-word phrases free of stop words which begin and end acceptably. Afterward, the most frequent phrases are selected as key word phrases.

Inventors

  • 👤 Francine R. Chen
  • 👤 Steven B. Putz
  • 👤 Daniel C. Brotsky

Bibliographic Details

Application Number —
Jurisdiction US
Data Source google_patents
Last Synchronized 2026-11-01 04:01
CPC Classifications
G; G06; G06F; G06F16/00; G06F16/30; G06F16/31; G06F16/313

Claims Analysis

Total Claims: 5
Claim 1
Independent Claim 361 words
An automatic method of selecting key phrases from a document presented in machine readable form to a processor, the document including a first multiplicity of words, some of the words forming phrases, the processor implementing the method by executing instructions stored in a memory device coupled to the processor, the method comprising the processor implemented steps of: a) generating from the document a multiplicity of phrases not including stop words; b) generating candidate phrases from the multiplicity of phrases, candidate phrases including more than one word and being composed of maximally long, non-overlapping subphrases, generating candidate phrases by the substeps of: b1) selecting a one of the multiplicity of phrases as a selected phrase; b2) selecting as a selected word a yet to be examined word of the selected phrase; b3) determining whether the selected word is a frequent word; b4) if the selected word is an infrequent: A) if all words of the selected phrase have not been examined, repeating steps b2) through b3); B) if all word of the selected phrase have been examined, repeating steps b1) through b3); C) if the selected word is a frequent word: C1) determining whether the selected word is an acceptable beginning for a key phrase by searching a list of bad beginning words; C2) if the selected word is not an acceptable beginning for a key phrase: i) determining whether all words of the selected phrase have been examined: ii) if all words of the selected phrase have not been examined, repeating steps b2) through b3); iii) if all words of the selected phrase have been examined, repeating steps b1) through b4); C3) if the selected word is an acceptable beginning for a key phrase: i) adding the selected word to a current phrase; ii) if all words of the selected phrase have not been examined selecting as a selected word a yet to be examined word of the selected phrase; iii) determining whether the selected word is a stop word; and iv) if the selected word is not a stop word, repeating steps C3i) through C3iii) c) selecting as key phrases a subset of most frequently occurring of the candidate phrases.
Claim 2
Dependent Claim 361 words
An automatic method of selecting key phrases from a document presented in machine readable form to a processor, the document including a first multiplicity of words, some of the words forming phrases, the processor implementing the method by executing instructions stored in a memory device coupled to the processor, the method comprising the processor implemented steps of: a) generating from the document a multiplicity of phrases not including stop words; b) generating candidate phrases from the multiplicity of phrases, candidate phrases including more than one word and being composed of maximally long, non-overlapping subphrases, generating candidate phrases by the substeps of: b1) selecting a one of the multiplicity of phrases as a selected phrase; b2) selecting as a selected word a yet to be examined word of the selected phrase; b3) determining whether the selected word is a frequent word; b4) if the selected word is an infrequent: A) if all words of the selected phrase have not been examined, repeating steps b2) through b3); B) if all word of the selected phrase have been examined, repeating steps b1) through b3); C) if the selected word is a frequent word: C1) determining whether the selected word is an acceptable beginning for a key phrase by searching a list of bad beginning words; C2) if the selected word is not an acceptable beginning for a key phrase: i) determining whether all words of the selected phrase have been examined: ii) if all words of the selected phrase have not been examined, repeating steps b2) through b3); iii) if all words of the selected phrase have been examined, repeating steps b1) through b4); C3) if the selected word is an acceptable beginning for a key phrase: i) adding the selected word to a current phrase; ii) if all words of the selected phrase have not been examined selecting as a selected word a yet to be examined word of the selected phrase; iii) determining whether the selected word is a stop word; and iv) if the selected word is not a stop word, repeating steps C3i) through C3iii) c) selecting as key phrases a subset of most frequently occurring of the candidate phrases.
Claim 3
Dependent Claim 135 words
The method of claim 1 wherein step C3) further comprises the steps of: v if the selected word is an infrequent word or if all words of the selected phrase have been examined: vA determining whether a last word of the current phrase is an acceptable ending for a key phrase by searching a list of bad ending words; vB if the last word of the current phrase is not an acceptable ending for a key phrase, removing the last word of the current phrase and repeating step vA); if the last word of the current phrase is an acceptable ending for a key phrase, determining whether the current phrase includes more than one word; and if the current phrase includes more than one word adding the current phrase to a list of candidate phrases.
Claim 4
Dependent Claim 55 words
The method of claim 1 wherein step a) further comprises: assigning a unique integer number to represent each word of the document; representing each candidate phrase as a string of integer numbers, each integer number of a string representing a word of the candidate phrase; and storing each string of integer numbers in a table.
Claim 5
Dependent Claim 51 words
The method of claim 3 wherein step a) further comprises: storing in a word table the unique integer number associated with each word of the document; for each word in the word table storing an indication of whether the word represents an acceptable beginning and an acceptable ending for candidate phrases.

References Cited (Backward Citations)

US4812966A Cited
US5553283A Cited
EP0361464A2 Cited
JPH02257266A Cited
JPH03105566A Cited
US5307266A Cited
US5251316A Cited
JPH0635961A Cited
US5440481A Cited
US5526443A Cited

USPTO Annuity & Maintenance Windows

Annuity Stage Window Opens Payment Deadline Surcharge Period Until Estimated Fee Status
3.5 Years (Year 4) 2001-04-27 2001-10-27 2002-04-28 $2000.00 Window Closed / Lapsed
7.5 Years (Year 8) 2005-04-27 2005-10-27 2006-04-28 $3760.00 Window Closed / Lapsed
11.5 Years (Year 12) 2009-04-27 2009-10-27 2010-04-28 $7700.00 Window Closed / Lapsed

Assigned Categories & Tags

No categories assigned to this patent yet.

Assign Category Tag