Method and System For Semantic Distance Measurement

• US 20100228733A1
• Filed: 11/11/2009
• Published: 09/09/2010
• Est. Priority Date: 11/12/2008
• Status: Active Grant

• Current Assignee: Zeta Global Corp. (Zeta Global Holdings Corp.)
• Original Assignee: Collective Media Incorporated
• Inventors: Harrison, Paul, Fulton, Hal, Roehrl, Armin, Oliphant, James
• Granted Patent: US 9,262,509 B2
• Time in Patent Office: Days
• US Class Current: 707/739
• CPC Class Codes: G06F 16/31 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/35 Clustering; Classification
• CPC Class Codes: G06F 16/36 Creation of semantic tools, e.g. ontology or thesauri

Abstract

A system and method for performing classification using semantic distance measurements. Items of electronic content accessed by individuals over a global communications network are identified. A set of content that includes the plurality of identified items of electronic content are stored. The set of content is normalized. Each of the keywords contained the set of content is identified and a semantic distance between each of the identified keywords is measured.

Claims

1. A computer readable storage medium having stored thereon computer executable instructions that, when executed on a computer, configure the computer to perform a method comprising:

identifying a plurality of items of electronic content accessed by a plurality of individuals over a global communications network;

storing a set of content comprising the plurality of identified items of electronic content;

normalizing the set of content;

identifying each of a plurality of keywords contained the set of content; and

measuring a semantic distance between each of the identified keywords.

2. The computer readable storage medium of claim 1, wherein measuring the semantic distance is calculated as 1−

τ

using the following formulas;

Keyword j has j not j totals Keyword i has i m₁₁ = n_ij m₁₂ = n_i −

n_ij m₁₊ = n_i not i m₂₁ = n_i −

n_ij m₂₂ = n −

n_i −

n_j + m₂₊ = m₂₁ + n_ij m₂₂ totals m₊₁ = n_i m₊₂ = m₁₂ + m₂₂ n

wherein

Keyword i is one of the identified keywords;

Keyword j is another one of the identified keywords;

a semantic distance between Keyword i and Keyword j is to be measured;

n is a total number of content items in the set;

n_i is a number of content items in the set that contain a keyword indexed by i;

n_j is a number of content items in the set that contain a keyword indexed by j; and

n_ij is a number of the content items containing both keywords indexed by i and j; and

$\mathrm{\tau <br><br>}=\frac{\frac{m_{11}^2}{n<br><br><br><br>m_{1+}}+\frac{m_{12}^2}{n<br><br><br><br>m_{1+}}+\frac{m_{21}^2}{n<br><br><br><br>m_{2+}}+\frac{m_{22}^2}{n<br><br><br><br>m_{2+}}-{\left(\frac{m_{+1}}{n}\right)}^2-{\left(\frac{m_{+2}}{n}\right)}^2}{1-{\left(\frac{m_{+1}}{n}\right)}^2-{\left(\frac{m_{+2}}{n}\right)}^2}.$

3. The computer readable storage medium of claim 1, the method further comprising:

classifying each of the plurality of identified items of electronic content in the set using, at least in part, the measured semantic distance.

4. The computer readable storage medium of claim 1 wherein the normalization is performed using a capture-recapture methodology.

5. A system comprising:

memory operable to store at least one program; and

at least one processor communicatively coupled to the memory, in which the at least one program, when executed by the at least one processor, causes the at least one processor to perform a method comprising the steps of;

identifying a plurality of items of electronic content accessed by a plurality of individuals over a global communications network;

storing a set of content comprising the plurality of identified items of electronic content;

normalizing the set of content;

identifying each of a plurality of keywords contained the set of content; and

measuring a semantic distance between each of the identified keywords.

6. The system of claim 5, wherein measuring the semantic distance is calculated as 1−

τ

using the following formulas;

Keyword j has j not j totals Keyword i has i m₁₁ = n_ij m₁₂ = n_i −

n_ij m₁₊= n_i not i m₂₁ = n_j −

n_ij m₂₂ = n −

n_i −

n_j + m₂₊ = m₂₁ + m₂₂ n_ij totals m₊₁ = n_i m₊₂ = m₁₂ + m₂₂ n

wherein

Keyword i is one of the identified keywords;

Keyword j is another one of the identified keywords;

a semantic distance between Keyword i and Keyword j is to be measured;

n is a total number of content items in the set;

n_i is a number of content items in the set that contain a keyword indexed by i;

n_j is a number of content items in the set that contain a keyword indexed by j; and

n_ij is a number of the content items containing both keywords indexed by i and j; and

$\mathrm{\tau <br><br>}=\frac{\frac{m_{11}^2}{n<br><br><br><br>m_{1+}}+\frac{m_{12}^2}{n<br><br><br><br>m_{1+}}+\frac{m_{21}^2}{n<br><br><br><br>m_{2+}}+\frac{m_{22}^2}{n<br><br><br><br>m_{2+}}-{\left(\frac{m_{+1}}{n}\right)}^2-{\left(\frac{m_{+2}}{n}\right)}^2}{1-{\left(\frac{m_{+1}}{n}\right)}^2-{\left(\frac{m_{+2}}{n}\right)}^2}.$

7. The system of claim 5, the method further comprising:

classifying each of the plurality of identified items of electronic content in the set using, at least in part, the measured semantic distance.

8. The system of claim 5 wherein the normalization is performed using a capture-recapture methodology.