Online marketing research utilizing sentiment analysis and tunable demographics analysis

• US 8,694,357 B2
• Filed: 06/07/2010
• Issued: 04/08/2014
• Est. Priority Date: 06/08/2009
• Status: Active Grant

• Current Assignee: Research Now Group LLC
• Original Assignee: e-Rewards
• Inventors: Ting, Tessie C., Davis, Jean, Pettit, Frances Annie
• Primary Examiner(s): Jarrett, Scott L
• Application Number:
  US12/802,445
• Publication Number:
  US 20110004483A1
• Time in Patent Office: 1,401 Days
• Field of Search: 705/7.29, 705/7.32
• US Class Current: 705/7.29
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data

Abstract

An online marketing research system where users identify a specific brand and/or competitive brands in which they are interested. An internet crawler engine collects sentiments relating to the identified brands according to a sampling method, which may be created by the user. The results from the internet crawler are refined using a refinement engine. The refined sentiments are then scored by a sentiment engine. Users may view the scored results via a user interface, which itself includes an interactive sentiment modeler. The interactive sentiment modeler provides quantified insights and allows users to select types of charts, the constructs, and timelines that are important to them.

Claims

1. A data processing system for providing targeted online marketing research comprising:

a sampling engine configured to perform stratified random sampling by stratifying an interne sampling frame;

a demographic boosting system, in communication with the sampling engine, configured to;

define target categories by selecting demographic variables for the interne sampling frame, wherein the demographic variables define a percentage of potential websites to be crawled that are associated with the demographic variables;

iteratively boost the demographic variables by assigning a weighted value to one or more of the demographic variables, and computing new target weights for each demographic variable by multiplying the average weighted value across all demographic variables by a default target weight assigned to the new target weight; and

tune the target categories by modifying searching parameters associated with potential Internet websites to be crawled, the modifiable search parameters comprising;

a list of potential interne websites to be crawled, a default target percentage of sentiments to be crawled for each potential internet website, and a specified percentage of a demographic variable for each of the potential interne websites;

a search engine, in communication with the demographic boosting system, configured to use the interne sampling frame to search and crawl internet website sentiments that are responsive to the target categories defined by the demographic boosting system;

a construct engine, in communication with the search engine, configured to store the interne website sentiments into taxonomic units of data; and

a sentiment modeler configured to present the website sentiments to a user in a plurality of charts, at least one of the charts being updateable to present updated website sentiments obtained in response to the iterative boosting of the demographic variables.

2. The data processing system as in claim 1 wherein the demographic variable associated with each of the potential websites includes:

a male variable defining a percentage of the potential interne websites to be crawled that are associated with males;

a female variable defining a percentage of the potential interne websites to be crawled that are associated with females;

an age variable defining a percentage of the potential internet websites to be crawled that are associated with a specified age range;

an income variable defining a percentage of the potential internet websites to be crawled that are associated with users having a specified financial income range; and

an education variable defining a percentage of the potential interne websites to be crawled that are associated with users having a specified education level.

3. The data processing system as in claim 2 wherein the demographic boosting system further includes computer readable instructions for:

enabling a demographic variable to be boosted by modifying any of the following variables;

the male variable, the female variable, the age variable, the income variable; and

the education variable;

computing, for each demographic variable, an average percentage for the demographic variable across all of the potential interne websites; and

computing, for each demographic variable, an average percentage for each of the potential interne websites.

4. The data processing system as in claim 3 wherein the demographic boosting system further includes computer readable instructions for boosting the demographic variables by assigning a weighted value to one or more of the demographic variables.

5. The data processing system as in claim 4 wherein if no boost is assigned to a demographic variable, assigning a weighted value of 100% to the demographic variable.

6. The data processing system as in claim 4 wherein the demographic boosting system further includes computer readable instructions for processing the demographic variable for each potential internet website by dividing the computed average demographic variable with the computed average percentage for the demographic variable across substantially all of the potential internet websites.

7. The data processing system as in claim 4 wherein the demographic boosting system further includes computer readable instructions for computing, for each potential internet website, an average weighted value across all of the demographic weights for the respective potential internet website.

8. The data processing system as in claim 1 wherein the new target weight reflects a percentage of sentiments to be pulled from each potential interne website.

9. The data processing system as in claim 1 wherein in determining internet website sentiments that are responsive to the target categories defined by the demographic boosting system, the search engine further including computer readable instructions for eliminating astro-turfing.

10. The data processing system as in claim 1 wherein the construct engine storing the sentiments into taxonomic units of data further includes computer readable instructions for:

identifying a client'"'"'s brand name;

defining the taxonomic units based on a pattern detected in the exploratory search results;

identifying keywords that are associated with each pattern;

processing a second search to confirm the exploratory search results; and

comparing the exploratory search results with the second search results.

11. The data processing system as in claim 1 wherein the taxonomic units of data are constructs providing average sentiment scores for sentiments using words relating to a product.

12. The data processing system as in claim 1 wherein the construct engine configured to store the sentiments into taxonomic units of data further includes computer readable instructions for scoring the sentiments based on a marketing mix of measures including price.

13. The data processing system as in claim 1 wherein the sampling engine configured to perform stratified random sampling further includes the sampling engine configured to create the internet sampling frame by crawling internet websites having a requisite level of internet traffic.

14. The data processing system as in claim 1 wherein the sampling engine configured to perform stratified random sampling further includes the sampling engine being configured to create the internet sampling frame using categorical stratified sampling including crawling internet websites having respective blogging content, respective microblogging content, respective image content, respective video content, respective social networking content, respective consumer ratings content, respective news content, and respective question and answer based content.

15. The data processing system as in claim 1 wherein the demographic boosting system, in communication with the sampling engine, defining target categories by selecting demographic variables for the internet sampling frame further includes selecting a matrix is based on demographics, verticals, and internet usage.

16. The data processing system as in claim 15 wherein the matrix is one of the following matrices:

an extensive contributor'"'"'s matrix targeting internet websites having content that is being constantly updated;

an extensive reader matrix targeting internet websites having a high volume of readers, regardless of the amount of contributors to the internet reader website;

a popular source matrix targeting internet websites having a high volume of registered and active users; and

a time sensitive matrix targeting internet websites having recently updated content.

17. The data processing system as in claim 15 wherein the matrix is one of the following matrices:

a financial matrix targeting internet websites having substantially high percentages of data about finances and money;

a business matrix targeting internet websites having substantially high percentages of data about business topics;

an apparel matrix targeting internet websites having substantially high percentages of data about clothing, shoes, and accessories; and

an electronics matrix targeting internet websites having substantially high percentages of data about electronics including televisions and music players.

18. The data processing system as in claim 15 wherein the matrix is one of the following matrices:

a sports matrix targeting internet websites having substantially high percentages of data about sporting goods and equipment;

an entertainment matrix targeting internet websites having substantially high percentages of data about current entertainment topics including movies and music;

a travel matrix targeting internet websites having substantially high percentages of data about traveling; and

a food and beverage matrix targeting internet websites having substantially high percentages of data about food and beverages.

19. The data processing system as in claim 15 wherein the matrix is one of the following matrices:

a restaurant matrix targeting internet websites having substantially high percentages of data about restaurants;

a medical matrix targeting internet websites having substantially high percentages of medical information;

a beauty matrix targeting internet websites having substantially high percentages of information about all types of beauty products; and

an automotive matrix targeting internet websites having substantially high percentages of information about vehicles.

20. The data processing system as in claim 15 wherein the matrix is one of the following matrices:

a home care matrix targeting internet websites having substantially high percentages of information about home care products;

a baby information matrix targeting internet websites having substantially high percentages of information about infants and toddlers aged 0 to 4;

a children information matrix targeting internet websites having substantially high percentages of information about children who are aged 5 to 12;

a teen contributors matrix targeting internet websites having substantially high percentages of users and readers who are aged 13 to 17; and

a teen information matrix targeting internet websites having substantially high percentages of data about people who are aged 13 to 17.

21. The data processing system as in claim 15 wherein the matrix is one of the following matrices:

an adult matrix targeting internet websites having substantially high percentages of users and readers who are aged 18 and older;

a male matrix targeting internet websites having substantially high percentages of users and readers who are male; and

a female matrix targeting internet websites having substantially high percentages of users and readers who are female.

22. The data processing system as in claim 15 wherein the matrix is one of the following matrices:

an affluent matrix targeting interne websites having substantially high percentages of users and readers who have incomes of $75,000 or more per year;

a low income matrix targeting internet websites having substantially high percentages of users and readers who have incomes less than $75,000 per year;

a scholars matrix targeting internet websites having substantially high percentages of users and readers who have at least a college degree; and

a low education matrix targeting internet websites having substantially high percentages of users and readers who do not have a college degree.

23. The data processing system as in claim 1 wherein selecting a matrix further includes defining a custom matrix by enabling the user to select a user defined website or user defined categories of websites.

24. A method for providing targeted online marketing research, the method comprising:

using a sampling engine to perform stratified random sampling on an internet sampling frame;

using a demographic boosting system, in communication with the sampling engine, to;

define target categories by selecting demographic variables for the internet sampling frame, wherein the demographic variables define a percentage of potential websites to be crawled that are associated with the demographic variables;

iteratively boost the demographic variables by assigning a weighted value to one or more of the demographic variables, and computing new target weights for each demographic variable by multiplying the average weighted value across all demographic variables by a default target weight assigned to the new target weight; and

tune the target categories by modifying searching parameters associated with potential internet websites to be crawled, the modifiable search parameters comprising;

a list of potential internet websites to be crawled, a default target percentage of sentiments to be crawled for each potential internet website, and a specified percentage of a demographic variable for each of the potential internet websites;

using a search engine, in communication with the demographic boosting system, to search and crawl internet website sentiments that are responsive to the target categories defined by the demographic boosting system;

using a construct engine, in communication with the search engine, to store the internet website sentiments into taxonomic units of data; and

using a sentiment modeler to present the website sentiments to a user in a plurality of charts, at least one of the charts being updateable to present updated website sentiments obtained in response to the iterative boosting of the demographic variables.

25. The method of claim 24 wherein the demographic variable associated with each of the potential websites includes:

a male variable defining a percentage of the potential internet websites to be crawled that are associated with males;

a female variable defining a percentage of the potential internet websites to be crawled that are associated with females;

an age variable defining a percentage of the potential internet websites to be crawled that are associated with a specified age range;

an income variable defining a percentage of the potential internet websites to be crawled that are associated with users having a specified financial income range; and

an education variable defining a percentage of the potential internet websites to be crawled that are associated with users having a specified education level.

26. The method of claim 25 wherein the demographic boosting system is further used to:

enable a demographic variable to be boosted by modifying any of the following variables;

the male variable, the female variable, the age variable, the income variable; and

the education variable;

compute, for each demographic variable, an average percentage for the demographic variable across all of the potential internet websites; and

compute, for each demographic variable, an average percentage for each of the potential internet websites.

27. The method of claim 26 wherein the demographic boosting system is further used to:

boost the demographic variables by assigning a weighted value to one or more of the demographic variables.

28. The method of claim 27 further comprising:

assigning a weighted value of 100% to the demographic variable if no boost is assigned to a demographic variable.

29. The method of claim 27 wherein the demographic boosting system is further used to:

process the demographic variable for each potential internet website by dividing the computed average demographic variable with the computed average percentage for the demographic variable across substantially all of the potential internet websites.

30. The method of claim 27 wherein the demographic boosting wherein the demographic boosting system is further used to:

compute, for each potential internet website, an average weighted value across all of the demographic weights for the respective potential internet website.

31. The method of claim 24 wherein the new target weight reflects a percentage of sentiments to be pulled from each potential internet website.

32. The method of claim 24 wherein in determining internet website sentiments that are responsive to the target categories defined by the demographic boosting system, the search engine is further used to eliminate astro-turfing.

33. The method of claim 24 wherein the construct engine storing the sentiments into taxonomic units of data is further used to:

identify a client'"'"'s brand name;

define the taxonomic units based on a pattern detected in the exploratory search results;

identify keywords that are associated with each pattern;

process a second search to confirm the exploratory search results; and

compare the exploratory search results with the second search results.

34. The method of claim 24 wherein the taxonomic units of data are constructs providing average sentiment scores for sentiments using words relating to a product.

35. The method of claim 24 wherein the construct engine configured to store the sentiments into taxonomic units of is further used to score the sentiments based on a marketing mix of measures including price.

36. The method of claim 24 wherein the sampling engine configured to perform stratified random sampling is further used to create the interne sampling frame by crawling internet websites having a requisite level of internet traffic.

37. The method of claim 24 wherein the sampling engine is further used to:

create the internet sampling frame using categorical stratified sampling, including crawling internet websites having respective blogging content, respective microblogging content, respective image content, respective video content, respective social networking content, respective consumer ratings content, respective news content, and respective question and answer based content.

38. The method of claim 24 wherein the demographic boosting system defines target categories by selecting demographic variables for the internet sampling frame by selecting a matrix based on demographics, verticals, and internet usage.

39. The method of claim 24 wherein selecting a matrix further includes defining a custom matrix by enabling the user to select a user defined website or user defined categories of websites.

40. A computer program product for providing targeted online marketing research, comprising:

a non-transitory computer-readable medium having program code recorded thereon, the program code comprising code to;

perform stratified random sampling on an interne sampling frame;

define target categories by selecting demographic variables for the interne sampling frame, wherein the demographic variables define a percentage of potential websites to be crawled that are associated with the demographic variables;

iteratively boost the demographic variables by assigning a weighted value to one or more of the demographic variables, and computing new target weights for each demographic variable by multiplying the average weighted value across all demographic variables by a default target weight assigned to the new target weight; and

tune the target categories by modifying searching parameters associated with potential internet websites to be crawled, the modifiable search parameters comprising;

a list of potential internet websites to be crawled, a default target percentage of sentiments to be crawled for each potential internet website, and a specified percentage of a demographic variable for each of the potential internet websites;

search and crawl internet website sentiments that are responsive to the target categories defined by the demographic boosting system;

store the internet website sentiments into taxonomic units of data; and

present the website sentiments to a user in a plurality of charts, at least one of the charts being updateable to present updated website sentiments obtained in response to the iterative boosting of the demographic variables.

41. The computer program product of claim 40 further comprising program code to:

define a percentage of the potential internet websites to be crawled that are associated with males;

define a percentage of the potential internet websites to be crawled that are associated with females;

define a percentage of the potential internet websites to be crawled that are associated with a specified age range;

define a percentage of the potential internet websites to be crawled that are associated with users having a specified financial income range; and

define a percentage of the potential internet websites to be crawled that are associated with users having a specified education level.

42. The computer program product of claim 41 further comprising program code to:

enable a demographic variable to be boosted by modifying any of the following variables;

the male variable, the female variable, the age variable, the income variable; and

the education variable;

compute, for each demographic variable, an average percentage for the demographic variable across all of the potential internet websites; and

compute, for each demographic variable, an average percentage for each of the potential internet websites.

43. The computer program product of claim 42 further comprising program code to:

boost the demographic variables by assigning a weighted value to one or more of the demographic variables.

44. The computer program product of claim 43 further comprising program code to:

assign a weighted value of 100% to the demographic variable if no boost is assigned to a demographic variable.

45. The computer program product of claim 43 further comprising program code to:

process the demographic variable for each potential internet website by dividing the computed average demographic variable with the computed average percentage for the demographic variable across substantially all of the potential internet websites.

46. The computer program product of claim 43 further comprising program code to:

compute, for each potential internet website, an average weighted value across all of the demographic weights for the respective potential internet website.

47. The computer program product of claim 40 further comprising program code to:

compute a percentage of sentiments to be pulled from each potential internet website.

48. The computer program product of claim 40 further comprising program code to:

determine internet website sentiments that are responsive to the target categories defined by a demographic boosting system.

49. The computer program product of claim 40 further comprising program code to:

identify a client'"'"'s brand name;

define the taxonomic units based on a pattern detected in the exploratory search results;

identify keywords that are associated with each pattern;

process a second search to confirm the exploratory search results; and

compare the exploratory search results with the second search results.

50. The computer program product of claim 40 further comprising program code to:

provide average sentiment scores for sentiments using words relating to a product.

51. The computer program product of claim 40 further comprising program code to:

score the sentiments based on a marketing mix of measures including price.

52. The computer program product of claim 40 further comprising program code to:

create an internet sampling frame by crawling internet websites having a requisite level of internet traffic.

53. The computer program product of claim 40 further comprising program code to:

create an internet sampling frame using categorical stratified sampling, including crawling internet websites having respective blogging content, respective microblogging content, respective image content, respective video content, respective social networking content, respective consumer ratings content, respective news content, and respective question and answer based content.

54. The computer program product of claim 40 further comprising program code to:

define target categories by selecting demographic variables for an internet sampling frame by selecting a matrix based on demographics, verticals, and internet usage.

55. The computer program product of claim 40 further comprising program code to:

define a custom matrix by enabling a user to select a user defined website or user defined categories of websites.