US 9,754,020 B1Granted

Method and device for measuring word pair relevancy

OwnerJon T. Kosloski, John W. Thompsoninventors

Life of this patent

Estimating the term…

Abstract

A method and device provide a word pair relevancy measure for a seed word-keyword pair. A set of documents is provided from which a co-occurrence matrix is created. The co-occurrence matrix is utilized to determine a transition probability and an expected search depth from which the word pair relevancy is determined.

Claims20

Dependent claims

A device for determining a measure of relevancy for a seed word-keyword pair (d,k) comprising:

a word identifier for identifying each unique word in the set of documents as a search word;
a word pair identifier in communication with said word identifier and for combining the identified search words to define search word pairs;
a unit portioner for portioning the set of documents into user-definable units, said unit portioner in communication with said interface and with said word pair identifier;
a co-occurrence matrix generator in communication with said unit portioner and said word pair identifier for determining, for each defined search word pair, the number of units in which the identified search word pair occurs and for storing the number of occurrences in a co-occurrence matrix
a probability matrix generator in communication with said co-occurrence matrix generator and for generating a probability matrix as a function of the co-occurrence matrix,
a calibrating column vector generator, ψ, in communication with said co-occurrence matrix generator for generating a calibrating column vector;
a matrix normalizer in communication with said probability matrix generator for normalizing the probability matrix to form a transition matrix, R;
a word pair selector in communication with said word pair identifier for selecting the seed word-keyword pair to be measured, wherein said word pair selector provides a first column vector, {right arrow over (Γ)}(d), relating to the seed word and a second column vector, {right arrow over (Γ)}(k), relating to the keyword;
an expected search distance generator in communication with said word pair selector, said probability matrix generator and said matrix normalizer, for calculating the expected search distance of the seed word-keyword pair;
a weighted average expected search distance generator in communication with said expected search distance generator, said probability matrix generator, said matrix normalizer, and said calibrating column vector, said weighted average expected search distance generator for determining a weighted average expected search distance for the keyword;
and
a calibrator in communication with said expected search distance generator and said weighted average expected search distance generator, wherein said calibrator determines the relevancy, sd,k, of the seed word to the key word, based upon said expected search distance and said weighted averaged expected search distance.

9 more elements

A method of determining a word pair relevancy measure for a seed word-keyword pair (d,k), comprising the steps of:

identifying a set of documents;
identifying search words from the set;
pairing a first search word and a second search word to define a search word pair;
portioning the set of documents into units;
creating a co-occurrence matrix identifying the number of co-occurrences of the search word pair in said units;
calculating a conditional co-occurrence probability associated with said search word pair, wherein said conditional co-occurrence probability is based upon said co-occurrence matrix;
calculating a transition probability associated with said search word pair, wherein said transition probability is based upon said co-occurrence matrix;
and
calculating an expected search distance for said seed word-keyword pair, wherein said expected search distance is based upon said conditional co-occurrence probability and said transition probability and wherein said expected search distance is defined as the expected number of units traversed to transition from the seed word to a unit that contains the keyword.

6 more elements

Dependencies are inferred from the claim text.

patentgrep

Loading patent…

Abstract

A device for determining a measure of relevancy for a seed word-keyword pair (d,k) comprising:

a word identifier for identifying each unique word in the set of documents as a search word;
a word pair identifier in communication with said word identifier and for combining the identified search words to define search word pairs;
a unit portioner for portioning the set of documents into user-definable units, said unit portioner in communication with said interface and with said word pair identifier;
a co-occurrence matrix generator in communication with said unit portioner and said word pair identifier for determining, for each defined search word pair, the number of units in which the identified search word pair occurs and for storing the number of occurrences in a co-occurrence matrix
a probability matrix generator in communication with said co-occurrence matrix generator and for generating a probability matrix as a function of the co-occurrence matrix,
a calibrating column vector generator, ψ, in communication with said co-occurrence matrix generator for generating a calibrating column vector;
a matrix normalizer in communication with said probability matrix generator for normalizing the probability matrix to form a transition matrix, R;
a word pair selector in communication with said word pair identifier for selecting the seed word-keyword pair to be measured, wherein said word pair selector provides a first column vector, {right arrow over (Γ)}(d), relating to the seed word and a second column vector, {right arrow over (Γ)}(k), relating to the keyword;
an expected search distance generator in communication with said word pair selector, said probability matrix generator and said matrix normalizer, for calculating the expected search distance of the seed word-keyword pair;
a weighted average expected search distance generator in communication with said expected search distance generator, said probability matrix generator, said matrix normalizer, and said calibrating column vector, said weighted average expected search distance generator for determining a weighted average expected search distance for the keyword;
and
a calibrator in communication with said expected search distance generator and said weighted average expected search distance generator, wherein said calibrator determines the relevancy, sd,k, of the seed word to the key word, based upon said expected search distance and said weighted averaged expected search distance.

9 more elements