Patent US85781A successfully refreshed from Google Patents.
EP0280866A2
Expired / Lapsed
EP
⏱️ 0.0 years remaining
Computer method for automatic extraction of commonly specified information from business correspondence
Assignee: International Business Machines Corp
•
Filing: 1988-01-22
•
Granted: —
•
Expires: 2008-01-22
Abstract
Abstract A Parametric Information Extraction (PIE) system has been developed to identify automatically commonly specified information such as author, date, recipient, address, subject statement, etc. from documents in free format. The program-generated data can be used directly or can be supplemented manually to provide automatic indexing or indexing aid, respectively. The PIE system uses structural, syntactic, and semantic knowledge to accomplish its objective. The structural analysis identifies the major components of a document which are the document heading, body, and ending. The heading and the ending which usually contain commonly specified information, are then analyzed by a battery of morphological, syntactic, and semantic pattern-matching procedures that provide specified information in standardized forms that can be easily manipulated by computer.
Inventors
- 👤 Elena Michelle Zamora
Bibliographic Details
Application Number
—
Jurisdiction
EP
Data Source
google_patents
Last Synchronized
2026-10-28 18:54
CPC Classifications
G; G06; G06F; G06F16/00; G06F16/30; G06F16/31; G06F16/313; G06F40/00; G06F40/20; G06F40/258
Claims Analysis
Total Claims: 7
Claim 1
Independent Claim
101 words
An information extraction method, for automatically identifying commonly specified information from documents in free format, comprising the steps of: reading in a document to be abstracted; reading in structural, syntactic and semantic knowledge data base identifying the major information components of the document using said structural knowledge data base;; analyzing the information components so identified, to obtain commonly specified information with said syntactic and said semantic knowledge data bases, employing pattern-matching procedures which provide commonly specified information in standardized form; outputting a formatted frame containing slots corresponding to the specified information that occur in the major information components of said document.
Claim 2
Dependent Claim
101 words
An information extraction method, for automatically identifying commonly specified information from documents in free format, comprising the steps of: reading in a document to be abstracted; reading in structural, syntactic and semantic knowledge data base identifying the major information components of the document using said structural knowledge data base;; analyzing the information components so identified, to obtain commonly specified information with said syntactic and said semantic knowledge data bases, employing pattern-matching procedures which provide commonly specified information in standardized form; outputting a formatted frame containing slots corresponding to the specified information that occur in the major information components of said document.
Claim 3
Dependent Claim
101 words
An automatic method of locating commonly specified information in a document in free format using pattern-matching techniques, comprising the steps of: identifying the date information in any location of a document using said pattern-matching techniques and a representation of the general syntactic pattern of dates; identifying personal names in any location of a document using said pattern-matching techniques and a specification of the constituents of personal names; identifying the subject of the document within the heading of the document by using said pattern-matching techniques and a specification of the formats used for titles, and other fields that contain the subject information.
Claim 4
Dependent Claim
101 words
An automatic method of locating commonly specified information in a document in free format using pattern-matching techniques, comprising the steps of: identifying the date information in any location of a document using said pattern-matching techniques and a representation of the general syntactic pattern of dates; identifying personal names in any location of a document using said pattern-matching techniques and a specification of the constituents of personal names; identifying the subject of the document within the heading of the document by using said pattern-matching techniques and a specification of the formats used for titles, and other fields that contain the subject information.
Claim 5
Dependent Claim
53 words
The method of claim 2, which further comprises: improving the relevance of information extracted from a document comprising the steps of: identification of the heading, body and ending of a document on the basis of syntactic, semantic and structural clues; isolation of the desired fields using only the relevant portions of the document.
Claim 6
Dependent Claim
73 words
An automatic method of mapping commonly specified information fields from a document in free format into standardized form comprising the steps of: identification of the semantic constituents of each field; application of pre-specified transformations to each constituent of said field and reformatting the resulting constituents in a prescribed order; placement of the standardized field in an output data structure which unambiguously identifies the type of data either by position or by explicit tagging.
Claim 7
Dependent Claim
73 words
An automatic method of mapping commonly specified information fields from a document in free format into standardized form comprising the steps of: identification of the semantic constituents of each field; application of pre-specified transformations to each constituent of said field and reformatting the resulting constituents in a prescribed order; placement of the standardized field in an output data structure which unambiguously identifies the type of data either by position or by explicit tagging.
References Cited (Backward Citations)
No citation records indexed.
USPTO Annuity & Maintenance Windows
Maintenance schedule is calculated for granted US utility patents.