Methods and systems for determining a risk of an emotional response of an audience

• US 10,521,807 B2
• Filed: 09/05/2013
• Issued: 12/31/2019
• Est. Priority Date: 09/05/2013
• Status: Active Grant

• Current Assignee: Tsg Technologies, LLC
• Original Assignee: Tsg Technologies, LLC
• Inventors: Sardella, Antonio
• Primary Examiner(s): Diaz, Susanna M.
• Application Number:
  US14/019,239
• Publication Number:
  US 20150066585A1
• Time in Patent Office: 2,308 Days
• Field of Search: None
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data

Abstract

A method for determining a risk of an emotional response of an audience to at least one publication is described. Additionally, a computing device for determining a risk of an emotional response of an audience to at least one publication is described. Moreover, a computer-readable storage device having processor-executable instructions embodied thereon, for determining a risk of an emotional response of an audience to at least one publication, is described.

Claims

1. A method for using a vector data structure to increase a speed of determining a risk of an emotional response of an audience to at least one publication, the method is implemented by a server computing device in communication with a database, the method comprising:

receiving, by the server computing device, a plurality of documents in an electronic format;

identifying, by the server computing device for each of a plurality of risk factors, a first set of the documents relevant to the risk factor and a second set of the documents not relevant to the risk factor, wherein each risk factor is a predetermined characteristic that influences the emotional response of the audience;

analyzing, by the server computing device for each of the risk factors, the corresponding first set of documents to determine occurrences and distributions of words within the first set of documents, and storing in the database, for each of the risk factors, an associated prototype vector, wherein each prototype vector includes a word set, each of the words in the word set weighted by a normalized frequency of occurrence of the word in the corresponding first set of documents;

comparing, by the server computing device, each prototype vector to occurrences and distributions of the words in the associated word set in the second set of documents not relevant to the associated risk factor to verify the prototype vector;

receiving, by the server computing device, the at least one publication, the at least one publication comprising one of a news item, an article, an opinion, a blog entry, and a social media message, the at least one publication communicated over a network, and converting the at least one publication from a first format to a second format by converting one of images, video, and audio to text that represents the at least one publication;

retrieving, by the server computing device from the database, the prototype vectors, each prototype vector stored in the database as a multi-dimensional data structure having a first column that includes each of the words in the word set, a second column that includes the normalized frequency of occurrence of each of the words in the word set, and a third column that includes the weight assigned to each of the words in the word set;

generating, by the server computing device, a test vector corresponding to each prototype vector, wherein each test vector includes each word in the corresponding prototype vector that also appears in the at least one publication;

determining, by the server computing device, a magnitude of each risk factor by comparing each test vector to the corresponding prototype vector;

retrieving, by the server computing device from the database, a model for weighting and summing the determined magnitudes of the risk factors, the model generated by the server computing device by performing principal component analysis (PCA) on a predetermined number of publications and performing an orthogonal transformation to determine a first principal component of the risk of the emotional response of the audience and a second principal component of the risk of the emotional response of the audience, and extracting eigenvalues from a correlation coefficient matrix to determine which risk factors most influence the audience;

determining, by the server computing device, a risk of an emotional response of the audience by weighting and summing the magnitudes of the risk factors according to the retrieved model, and transmitting the risk of the emotional response of the audience to at least one client computing device; and

generating based on the risk of the emotional response before an emotional response occurs and outputting, by the server computing device to a user interface displayed on a display of the at least one client computing device, an electronic user interactive visual representation of the risk of the emotional response in a first geographic region and a second geographic region that is different from the first geographic region that simultaneously shows a variation of the risk of the emotional response in the first geographic region and the second geographic region that is different from the first geographic region, the electronic user interactive visual representation showing on the display for both the first geographic region and the second geographic region a respective point on a graph of the first principal component of the risk of the emotional response of the audience having a range from negative one representing a low risk to positive one representing a high risk along a first axis and a second principal component of the risk of the emotional response of the audience having a range from negative one representing the low risk to positive one representing the high risk along a second axis.

2. The method of claim 1, wherein:

receiving at least one publication further comprises;

receiving an identification of a first topic; and

receiving at least one publication that pertains to the first topic; and

determining a risk of an emotional response further comprises determining a risk of an emotional response pertaining to the first topic.

3. The method of claim 1, wherein receiving at least one publication further comprises:

receiving an identification of a first publisher; and

receiving the at least one publication from the first publisher.

4. The method of claim 1, further comprising:

generating a first set of prototype vectors for a first audience; and

generating a second set of prototype vectors for a second audience, wherein the first set of prototype vectors is different from the second set of prototype vectors.

5. The method of claim 4, wherein:

generating a first set of prototype vectors for a first audience further comprises generating a first set of prototype vectors for a first audience that is located in the first geographic region; and

generating a second set of prototype vectors for a second audience further comprises generating a second set of prototype vectors for a second audience that is located in the second geographic region that is different from the first geographic region.

6. The method of claim 1, wherein retrieving the prototype vectors further comprises determining a category of the audience and retrieving a set of prototype vectors associated with the category.

7. The method of claim 1, wherein retrieving the model further comprises determining a category associated with the audience and retrieving a model associated with the category.

8. The method of claim 1, wherein generating a test vector corresponding to each prototype vector further includes:

indexing the at least one received publication; and

weighting each word in each test vector based at least on a frequency of the word in the at least one indexed publication.

9. The method of claim 1, further comprising generating the model for weighting and summing the magnitudes of the risk factors by performing the principal component analysis.

10. The method of claim 1, further comprising:

associating, by the computing device in the database, the at least one publication as at least one historical publication of a first publisher;

receiving, by the computing device, at least one new publication associated with the first publisher;

generating, by the computing device, a new test vector corresponding to each prototype vector, wherein each new test vector includes each word in the corresponding prototype vector that also appears in the at least one new publication;

determining, by the computing device, a magnitude of each risk factor associated with the at least one new publication by comparing each new test vector to the corresponding prototype vector;

determining, by the computing device, a change in the magnitude of at least one risk factor associated with the at least one new publication as compared to the magnitude of the corresponding at least one risk factor associated with the at least one historical publication; and

outputting, by the computing device to at least one client computing device, an alert identifying the change in usage of the at least one risk factor by the first publisher.

11. A server computing device for using a vector data structure to increase a speed of determining a risk of an emotional response of an audience to at least one publication, the server computing device is communicatively coupled to a database, the server computing device comprising:

a processor; and

a memory device in communication with the processor, the processor to;

receive a plurality of documents in an electronic format;

identify, for each of a plurality of risk factors, a first set of the documents relevant to the risk factor and a second set of the documents not relevant to the risk factor, wherein each risk factor is a predetermined characteristic that influences the emotional response of the audience;

analyze, for each of the risk factors, the corresponding first set of documents to determine occurrences and distributions of words within the first set of documents, and store in the database, for each of the risk factors, an associated prototype vector, wherein each prototype vector includes a word set, each of the words in the word set weighted by a normalized frequency of occurrence of the word in the corresponding first set of documents;

compare each prototype vector to occurrences and distributions of the words in the associated word set in the second set of documents not relevant to the associated risk factor to verify the prototype vector;

receive the at least one publication, the at least one publication comprising one of a news item, an article, an opinion, a blog entry, and a social media message, the at least one publication communicated over a network, and convert the at least one publication from a first format to a second format by converting one of images, video, and audio to text that represents the at least one publication;

retrieve, from the database, the prototype vectors, each prototype vector stored in the database as a multi-dimensional data structure having a first column that includes each of the words in the word set, a second column that includes the normalized frequency of occurrence of each of the words in the word set, and a third column that includes the weight assigned to each of the words in the word set;

generate a test vector corresponding to each prototype vector, wherein each test vector includes each word in the corresponding prototype vector that also appears in the at least one publication;

determine a magnitude of each risk factor by comparing each test vector to the corresponding prototype vector;

retrieve, from the database, a model for weighting and summing the determined magnitudes of the risk factors, the model generated by the server computing device by performing principal component analysis (PCA) on a predetermined number of publications and performing an orthogonal transformation to determine a first principal component of the risk of the emotional response of the audience and a second principal component of the risk of the emotional response of the audience, and extracting eigenvalues from a correlation coefficient matrix to determine which risk factors most influence the audience;

determine a risk of an emotional response of the audience by weighting and summing the magnitudes of the risk factors according to the retrieved model, and transmitting the risk of the emotional response of the audience to at least one client computing device; and

generate based on the risk of the emotional response before an emotional response occurs and output to a user interface displayed on a display of the at least one client computing device an electronic user interactive graphical representation of the risk of the emotional response in a first geographic region and a second geographic region that is different from the first geographic region that simultaneously shows a variation of the risk of the emotional response in the first geographic region and the second geographic region that is different from the first geographic region, the electronic user interactive visual representation showing on the display for both the first geographic region and the second geographic region a respective point on a graph of the first principal component of the risk of the emotional response of the audience having a range from negative one representing a low risk to positive one representing a high risk along a first axis and a second principal component of the risk of the emotional response of the audience having a range from negative one representing the low risk to positive one representing the high risk along a second axis.

12. The computing device of claim 11, wherein the processor is further to retrieve the prototype vectors by determining a category associated with the audience and retrieving a set of prototype vectors associated with the category.

13. The computing device of claim 11, wherein the processor is further to retrieve the model by determining a category associated with the audience and retrieving a model associated with the category.

14. The computing device of claim 11, wherein the processor is further to generate a test vector corresponding to each prototype vector by:

indexing the at least one received publication; and

weighting each word in each test vector based on at least a frequency of the word in the at least one indexed publication.

15. The computing device of claim 11, wherein the processor is further to generate the model for weighting and summing the magnitudes of the risk factors by performing the principal component analysis.

16. The computing device of claim 11, wherein the processor is further to:

associate, in the database, the at least one publication as at least one historical publication of a first publisher;

receive at least one new publication associated with the first publisher;

generate a new test vector corresponding to each prototype vector, wherein each new test vector includes each word in the corresponding prototype vector that also appears in the at least one new publication;

determine a magnitude of each risk factor associated with the at least one new publication by comparing each new test vector to the corresponding prototype vector;

determine a change in the magnitude of at least one risk factor associated with the at least one new publication as compared to the magnitude of the corresponding at least one risk factor associated with the at least one historical publication; and

output to at least one client computing device an alert identifying the change in usage of the at least one risk factor by the first publisher.

17. A non-transitory computer-readable storage device having processor-executable instructions embodied thereon for using a vector data structure to increase a speed of determining a risk of an emotional response of an audience to at least one publication, wherein when executed by a server computing device communicatively coupled to a database, the processor-executable instructions cause the server computing device to:

receive a plurality of documents in an electronic format;

identify, for each of a plurality of risk factors, a first set of the documents relevant to the risk factor and a second set of the documents not relevant to the risk factor, wherein each risk factor is a predetermined characteristic that influences the emotional response of the audience;

analyze, for each of the risk factors, the corresponding first set of documents to determine occurrences and distributions of words within the first set of documents, and store in the database, for each of the risk factors, an associated prototype vector, wherein each prototype vector includes a word set, each of the words in the word set weighted by a normalized frequency of occurrence of the word in the corresponding first set of documents;

compare each prototype vector to occurrences and distributions of the words in the associated word set in the second set of documents not relevant to the associated risk factor to verify the prototype vector;

receive the at least one publication, the at least one publication comprising one of a news item, an article, an opinion, a blog entry, and a social media message, the at least one publication communicated over a network, and convert the at least one publication from a first format to a second format by converting one of images, video, and audio to text that represents the at least one publication;

retrieve, from the database, the prototype vectors, each prototype vector stored in the database as a multi-dimensional data structure having a first column that includes each of the words in the word set, a second column that includes the normalized frequency of occurrence of each of the words in the word set, and a third column that includes the weight assigned to each of the words in the word set;

generate a test vector corresponding to each prototype vector, wherein each test vector includes each word in the corresponding prototype vector that also appears in the at least one publication;

determine a magnitude of each risk factor by comparing each test vector to the corresponding prototype vector;

retrieve, from the database, a model for weighting and summing the determined magnitudes of the risk factors, the model generated by the server computing device by performing principal component analysis (PCA) on a predetermined number of publications and performing an orthogonal transformation to determine a first principal component of the risk of the emotional response of the audience and a second principal component of the risk of the emotional response of the audience, and extracting eigenvalues from a correlation coefficient matrix to determine which risk factors most influence the audience;

determine a risk of an emotional response of the audience by weighting and summing the magnitudes of the risk factors according to the retrieved model, and transmit the risk of the emotional response of the audience to at least one client computing device; and

generate based on the risk of the emotional response before an emotional response occurs and output to a user interface displayed on a display of the at least one client computing device, an electronic user interactive graphical representation of the risk of the emotional response in a first geographic region and a second geographic region that is different from the first geographic region that simultaneously shows a variation of the risk of the emotional response in the first geographic region and the second geographic region that is different from the first geographic region, the electronic user interactive visual representation showing on the display for both the first geographic region and the second geographic region a respective point on a graph of the first principal component of the risk of the emotional response of the audience having a range from negative one representing a low risk to positive one representing a high risk along a first axis and a second principal component of the risk of the emotional response of the audience having a range from negative one representing the low risk to positive one representing the high risk along a second axis.

18. The non-transitory computer-readable storage device of claim 17, wherein the processor-executable instructions further cause the computing device to:

associate, in the database, the at least one publication as at least one historical publication of a first publisher;

receive at least one new publication associated with the first publisher;

generate a new test vector corresponding to each prototype vector, wherein each new test vector includes each word in the corresponding prototype vector that also appears in the at least one new publication;

determine a magnitude of each risk factor associated with the at least one new publication by comparing each new test vector to the corresponding prototype vector;

determine a change in the magnitude of at least one risk factor associated with the at least one new publication as compared to the magnitude of the corresponding at least one risk factor associated with the at least one historical publication; and

output to at least one client computing device an alert identifying the change in usage of the at least one risk factor by the first publisher.