Granted Patent B2

US 9,922,352 · App. 15/005,840 · Granted Mar 20, 2018

Abstract

Claims (38)

1. A method, implemented by one or more processors in a computing system, for generating a multidimensional synopsis of a stream of textual data, the method comprising:

accessing, by the one or more processors, a stream of textual data that includes a number of elements of textual data, each element of textual data comprising plain text content that is associated with an author and is directed to a particular subject;

identifying, by the one or more processors, a first dimension and a second dimension for the stream of textual data, the first dimension including a number of concepts that each represent a subject attribute of the particular subject, the second dimension including a number of concepts that each represent an author attribute;

processing, by the one or more processors, each of the number of elements of textual data to identify which of the concepts of the first and second dimension appear in the plain text content included in the element, and for each concept within the first dimension that appears in the plain text content included in the element, generating a quantitative value; and

generating, by the one or more processors, the multidimensional synopsis of the stream of textual data by generating a score for each intersecting set of concepts from the corresponding quantitative values, each score representing a prevalence of the intersecting set of concepts within the stream of textual data.

2. The method of claim 1 , wherein the quantitative value defines a sentiment of the author of the plain text content included in the element of textual data towards the subject attribute represented by the concept.

3. The method of claim 1 , wherein the quantitative value defines an occurrence of a question directed towards the subject attribute represented by the concept.

4. The method of claim 1 , wherein the score is generated by summing the quantitative values.

5. The method of claim 1 , wherein the score for each intersecting set of concepts includes a positive component and a negative component.

6. The method of claim 1 , wherein the first dimension and second dimension and the concepts of each dimension are generated by analyzing the stream of textual data.

7. The method of claim 1 , wherein processing each of the number of elements of textual data to identify which of the concepts of the first and second dimension appear in the plain text content included in the element comprises performing natural language processing on the number of elements of textual data.

8. The method of claim 1 , wherein identifying a first dimension and a second dimension for the stream of textual data further includes identifying one or more additional dimensions, each additional dimension including a number of concepts;

wherein processing each of the number of elements of textual data to identify which of the concepts of the first and second dimension appear in the plain text content included in the element further includes processing each of the number of elements of textual data to identify which of the concepts of each of the one or more additional dimension appear in the plain text content included in the element; and

wherein each intersecting set of concepts includes a concept from at least two of the dimensions.

9. The method of claim 1 , wherein the elements of textual data comprise user reviews of a product such that the first dimension includes concepts that represent attributes of the product and the second dimension includes concepts that represent possible classifications of users.

10. One or more computer storage media storing computer executable instructions which when executed by one or more processors implements a method for generating a multidimensional synopsis of a stream of textual data, the method comprising:

accessing a stream of textual data that includes a number of elements of textual data, each element of textual data comprising plain text content that is associated with an author and is directed to a particular subject;

identifying a first dimension and a second dimension for the stream of textual data, the first dimension including a number of concepts that each represent a subject attribute of the particular subject, the second dimension including a number of concepts that each represent an author attribute;

generating machine learning classification training for the concepts in the first and second dimensions;

for each of the number of elements of textual data, processing the element against the machine learning classification training to identify which concepts appear in the plain text content included in the element, and for each concept within the first dimension that appears in the plain text content included in the element, generating a quantitative value;

identifying each intersecting set of concepts from the first and second dimensions; and

for each intersecting set of concepts, generating a score from the corresponding quantitative values, the score representing a prevalence of the intersecting set of concepts within the stream of textual data.

11. The computer storage media of claim 10 , wherein the quantitative values are sentiment values.

12. The computer storage media of claim 10 , wherein the score for each intersecting set of concepts includes a positive component and a negative component.

13. The computer storage media of claim 10 , wherein identifying a first dimension and a second dimension for the stream of textual data further includes identifying one or more additional dimensions, each additional dimension including a number of concepts; and

wherein identifying each intersecting set of concepts from the first and second dimensions comprises identifying at least some intersecting sets of concepts from the first, second, and one or more additional dimensions.

14. A system comprising:

one or more processors; and

computer storage media storing computer executable instructions which when executed perform a method for generating a multidimensional synopsis of a stream of textual data, the method comprising:

accessing a stream of textual data that includes a number of elements of textual data, each element of textual data comprising plain text content that is associated with an author and is directed to a particular subject;

identifying a first dimension and a second dimension for the stream of textual data, the first dimension including a number of concepts that each represent a subject attribute of the particular subject, the second dimension including a number of concepts that each represent an author attribute;

generating machine learning classification training for the concepts in the first and second dimensions;

for each of the number of elements of textual data, determining, using the machine learning classification training, which sentence fragments within the plain text content included in the element address a particular concept of the first or second dimension, and for each concept within the first dimension that appears in the plain text content included in the element, generating a quantitative value;

identifying each intersecting set of concepts from the first and second dimensions; and

for each intersecting set of concepts, generating a score from the corresponding quantitative values, the score representing a prevalence of the intersecting set of concepts within the stream of textual data.

15. The system of claim 14 , wherein each quantitative value defines one of:

a sentiment of the author of the plain text content included in the element of textual data towards the subject attribute represented by the concept; or

an occurrence of a question directed towards the subject attribute represented by the concept.

Assignments (32)