System and method for adapting sentiment analysis to user profiles to reduce bias
Provided is a system and method for adapting sentiment analysis to user profiles to reduce bias in customer or user generated content, specifically a system and method that discounts or adjusts bias in sentiment data based on the channel from which the content was received and/or the demographic of the user. The system includes a means to detect sentiment bias for any product, service, or company across multiple channels of customer data; a means to construct models to quantize bias by specific demographics and channels; and a means to adjust sentiment model output to reduce inflation by biased groups.
1. A method of improving objectivity of an outcome of regression analysis across a plurality of customer service channels, wherein each customer service channel is an electronic platform, the method including one or more processing devices performing operations comprising:
building demographic profiles for users in interactions across the plurality of customer service channels, wherein the customer service channel is chat, email, telephonic, or a social media platform;
grouping all of the interactions around a common product, topic or service to produce at least one grouping of interactions;
performing sentiment analysis on content of each grouping by each user on each customer service channel based on the demographic profiles and each customer service channel to quantize bias in a segment of the content by specific demographic and channel to produce an original sentiment score for each grouping;
performing a regression analysis on the content of each grouping to determine where sentiment bias resides for a particular combination of user demographic and customer service channel and constructing a model of correlations between specific customer attributes, customer service channels, and sentiment polarity according to the regression analysis for each grouping;
determining a sentiment adjustment factor based on the correlations; and
applying the sentiment adjustment factor to the original sentiment score to compensate for the sentiment bias in feedback obtained from the particular combination of user demographic and customer service channel for the grouping.
2. The method of claim 1 , wherein the regression analysis is performed on content created on each customer service channel by each user.
3. The method of claim 1 , wherein the regression analysis is performed in parallel via a distributed computer cluster.
4. The method of claim 3 , wherein the parallel performing of the regression analysis is disturbed according to each customer service channel.
5. The method of claim 1 , further comprising measuring via the sentiment analysis a tendency for a certain demographic to have a more positive or negative sentiment than a population of users as a whole and wherein the sentiment adjustment factor measures take into account this tendency.
6. The method of claim 1 , further comprising measuring via the sentiment analysis a tendency for communication via a certain customer service channel to have a more positive or negative sentiment than customer service channels as a whole and wherein the sentiment adjustment factor measures take into account this tendency.
7. A system improving objectivity of an outcome of regression analysis across a plurality of customer service channels, wherein each customer service channel is an electronic platform, comprising:
a memory comprising executable instructions; and
a processor configured to execute the executable instructions and cause the system to:
building demographic profiles for users in interactions across the plurality of customer service channels, wherein the customer service channel is chat, email, telephonic, or a social media platform;
group all of the interactions around a common product, topic or service to produce at least one grouping of interactions;
perform sentiment analysis on content of each grouping by each user on each customer service channel based on the demographic profiles and each customer service channel to quantize bias in a segment of the content by specific demographic and channel to produce an original sentiment score for each grouping;
perform a regression analysis on the content of each grouping to determine where sentiment bias resides for a particular combination of user demographic and user channel and construct a model of correlations between specific customer attributes, customer service channels, and sentiment polarity according to the regression analysis for each grouping;
determine a sentiment adjustment factor based on the correlations; and
apply the sentiment adjustment factor to the original sentiment score to compensate for the sentiment bias in feedback obtained from the particular combination of user demographic and customer service channel for the grouping.
8. The system of claim 7 , wherein the regression analysis is performed on content created on each customer service channel by each user.
9. The system of claim 7 , wherein the regression analysis is performed in parallel via a distributed computer cluster.
10. The system of claim 9 , wherein the parallel performing of the regression analysis is disturbed according to each customer service channel.
11. The system of claim 7 , the memory further comprising executable instructions, that cause the system to measure, via the sentiment analysis, a tendency for a certain demographic to have a more positive or negative sentiment than a population of users as a whole and wherein the sentiment adjustment factor measures take into account this tendency.
12. The system of claim 7 , the memory further comprising the memory further comprising executable instructions, that cause the system to measure, via the sentiment analysis, a tendency for communication via a certain customer service channel to have a more positive or negative sentiment than customer service channels as a whole and wherein the sentiment adjustment factor measures take into account this tendency.
13. A non-transitory computer-readable storage medium comprising instructions that, when executed by a processor of a processing system, cause the processing system to perform a method, the method comprising:
building demographic profiles for users in interactions across a plurality of customer service channels, wherein the customer service channel is chat, email, telephonic, or a social media platform;
grouping all of the interactions around a common product, topic or service to produce at least one grouping of interactions;
performing sentiment analysis on content of each grouping by each user on each customer service channel based on the demographic profiles and each customer service channel to quantize bias in a segment of the content by specific demographic and channel to produce an original sentiment score for each grouping;
performing a regression analysis on the content of each grouping to determine where sentiment bias resides for a particular combination of user demographic and customer service channel and constructing a model of correlations between specific customer attributes, customer service channels, and sentiment polarity according to the regression analysis for each grouping;
determining a sentiment adjustment factor based on the correlations; and;
applying the sentiment adjustment factor to the original sentiment score to compensate for the sentiment bias in feedback obtained from the particular combination of user demographic and customer service channel for the grouping.
14. The non-transitory computer-readable storage medium of claim 13 , wherein the regression analysis is performed on content created on each customer service channel by each user.
15. The non-transitory computer-readable storage medium of claim 13 , wherein the regression analysis is performed in parallel via a distributed computer cluster.
16. The non-transitory computer-readable storage medium of claim 15 , wherein the parallel performing of the regression analysis is disturbed according to each customer service channel.
17. The non-transitory computer-readable storage medium of claim 13 , the method further comprising measuring via the sentiment analysis a tendency for a certain demographic to have a more positive or negative sentiment than a population of users as a whole and wherein the sentiment adjustment factor measures take into account this tendency.
18. The non-transitory computer-readable storage medium of claim 13 , the method further comprising measuring via the sentiment analysis a tendency for communication via a certain customer service channel to have a more positive or negative sentiment than customer service channels as a whole and wherein the sentiment adjustment factor measures take into account this tendency.