System, method and computer program product for identifying products associated with polarized sentiments

• US 8,600,796 B1
• Filed: 01/30/2012
• Issued: 12/03/2013
• Est. Priority Date: 01/30/2012
• Status: Active Grant

• Current Assignee: Bazaarvoice Incorporated
• Original Assignee: Bazaarvoice Incorporated
• Inventors: Sterne, Brendan, Friesenhahn, Dustin, Sedota, Donald J. Jr.
• Primary Examiner(s): Boyce, Andre
• Assistant Examiner(s): FEACHER, LORENA R
• Application Number:
  US13/361,691
• Time in Patent Office: 673 Days
• Field of Search: 705/7.29, 705 731- 734
• US Class Current: 705/7.33
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G06Q 30/0203 Market surveys; Market polls

Abstract

An overall average review rating for a product may be determined, based on user ratings that are associated with opinions of a product, within a dimension corresponding to a user trait. A segment variation score for each of a plurality of segments of the dimension may be determined. Each segment may correspond to one or more values of the user trait corresponding to the dimension. A total variation score may be determined for the dimension based on the segment variation scores determined for each of the plurality of segments of the dimension. The total variation score for the dimension may be compared to a polarization threshold to determine whether the dimension is polarized. Generated information may identify the product as having sentiment that is polarized.

Claims

1. A method for identifying products with polarized sentiments, the method comprising:

obtaining, by a computer system, a plurality of user ratings associated with a product, wherein each user rating corresponds to a respective user'"'"'s opinion of the product;

determining, by the computer system, an overall average review rating for the product within a dimension, wherein the dimension corresponds to a user trait, and wherein the determining the overall average review rating is based on the user ratings;

determining, by the computer system, a segment variation score for each of a plurality of segments of the dimension, wherein each segment corresponds to one or more values of the user trait corresponding to the dimension, and wherein the determining the segment variation score for a particular segment is based on a distance between the overall average review rating and an average of a portion of the user ratings corresponding to the particular segment;

determining, by the computer system, a total variation score for the dimension based on the segment variation scores determined for each of the plurality of segments of the dimension;

determining, by the computer system, whether the dimension is polarized by comparing the total variation score for the dimension to a polarization threshold; and

in response to determining that the dimension is polarized, the computer system generating information configured to identify the product as having sentiment that is polarized with respect to the dimension.

2. The method of claim 1, further comprising:

determining the portion of the user ratings corresponding to the particular segment, wherein the portion corresponds to opinions of users having a particular value for the trait corresponding to the dimension.

3. The method of claim 1, further comprising:

determining a p-value for the dimension based on a distribution of sentiment variation in the dimension, wherein the determining whether the dimension is polarized includes comparing the determined p-value to a p-value threshold.

4. The method of claim 1,

wherein the user trait corresponds to user'"'"'s gender;

and

wherein the plurality of segments includes;

a segment corresponding to male users; and

a segment corresponding to female users.

5. The method of claim 1, wherein the generated information is configured to cause graphical display of the sentiment being polarized with respect to the dimension.

6. The method of claim 5, further comprising sending the generated information to a device that is different from the computer system.

7. A system, comprising:

at least one processor;

at least one non-transitory computer readable medium storing instructions executable by the system, using the at least one processor, to implement a product polarization module, the product polarization module being configured for;

obtaining a plurality of user ratings associated with a product, wherein each user rating corresponds to a respective user'"'"'s opinion of the product;

determining an overall average review rating for the product within a dimension, wherein the dimension corresponds to a user trait, and wherein the determining the overall average review rating is based on the user ratings;

determining a segment variation score for each of a plurality of segments of the dimension, wherein each segment corresponds to one or more values of the user trait corresponding to the dimension, and wherein the determining the segment variation score for a particular segment is based on a distance between the overall average review rating and an average of a portion of the user ratings corresponding to the particular segment;

determining a total variation score for the dimension based on the segment variation scores determined for each of the plurality of segments of the dimension;

determining whether the dimension is polarized by comparing the total variation score for the dimension to a polarization threshold; and

in response to determining that the dimension is polarized, generating information configured to identify the product as having sentiment that is polarized with respect to the dimension.

8. The system of claim 7, wherein the product polarization module is further configured for:

determining the portion of the user ratings corresponding to the particular segment, wherein the portion corresponds to opinions of users having a particular value for the trait corresponding to the dimension.

9. The system of claim 7, wherein the product polarization module is further configured for:

determining a p-value for the dimension based on a distribution of sentiment variation in the dimension, wherein the determining whether the dimension is polarized includes comparing the determined p-value to a p-value threshold.

10. The system of claim 7,

wherein the user trait corresponds to user'"'"'s age;

and

wherein the plurality of segments includes;

a first segment corresponding to users having ages within a first age range; and

a second segment corresponding to users having ages within a second age range.

11. The system of claim 7, wherein the generated information is configured to cause graphical display of the sentiment being polarized with respect to the dimension.

12. The system of claim 11, wherein the product polarization module is further configured for sending the generated information to a device that is different from the computer system.

13. An article of manufacture comprising a non-transitory computer-readable medium having stored thereon instructions executable by a computer system to cause the computer system to perform operations comprising:

obtaining a plurality of user ratings associated with a product, wherein each user rating corresponds to a respective user'"'"'s opinion of the product;

determining an overall average review rating for the product within a dimension, wherein the dimension corresponds to a user trait, and wherein the determining the overall average review rating is based on the user ratings;

determining a segment variation score for each of a plurality of segments of the dimension, wherein each segment corresponds to one or more values of the user trait corresponding to the dimension, and wherein the determining the segment variation score for a particular segment is based on a distance between the overall average review rating and an average of a portion of the user ratings corresponding to the particular segment;

determining a total variation score for the dimension based on the segment variation scores determined for each of the plurality of segments of the dimension;

determining whether the dimension is polarized by comparing the total variation score for the dimension to a polarization threshold; and

in response to determining that the dimension is polarized, generating information configured to identify the product as having sentiment that is polarized with respect to the dimension.

14. The article of manufacture of claim 13, wherein the operations further comprise:

determining the portion of the user ratings corresponding to the particular segment, wherein the portion corresponds to opinions of users having a particular value for the trait corresponding to the dimension.

15. The article of manufacture of claim 13, wherein the operations further comprise:

determining a p-value for the dimension based on a distribution of sentiment variation in the dimension, wherein the determining whether the dimension is polarized includes comparing the determined p-value to a p-value threshold.

16. The article of manufacture of claim 13,

wherein the user trait corresponds to user'"'"'s marital status;

and

wherein the plurality of segments includes;

a segment corresponding to single users; and

a segment corresponding to married users.

17. The article of manufacture of claim 13,

wherein the user trait corresponds to user'"'"'s method of purchasing the product;

and

wherein the plurality of segments includes a segment corresponding to users that purchased the products online.

18. The article of manufacture of claim 13,

wherein the user trait corresponds to user'"'"'s income.

19. The article of manufacture of claim 13, wherein the generated information is configured to cause graphical display of the sentiment being polarized with respect to the dimension.

20. The article of manufacture of claim 19, wherein the operations further comprise sending the generated information to a device that is different from the computer system.