Granted Patent B2

US 9,299,081 · App. 13/608,149 · Granted Mar 29, 2016

Abstract

Claims (59)

1. A method for deriving a user profile, comprising:

using a processor device, performing steps of:

generating a user interface on which a user, for whom no profile exists, interacts in a session;

formulating questions such that a user response to the questions indicates likes and dislikes of the user;

initially presenting the questions representing topics from different high-level categories to the user on the user interface;

receiving responses from the user, the responses indicating user interest from among the different high-level categories;

iteratively performing the following for each high-level category in which the user expressed an interest:

selecting a next level of cluster of questions from an increasingly narrowed-down subset of clusters, when one or more responses to questions from a previous level of cluster indicates user interest;

wherein all questions in a cluster pertain to a same topic;

presenting questions from the selected next level of cluster;

receiving user responses to the questions; and

deriving user preferences from the user responses;

wherein the iterative acts are performed repeatedly until termination of the session;

inhibiting at least one cluster, including one or more sub-clusters within the at least one cluster, from consideration for the session when the user indicates no interest in at least one topic associated with the at least one cluster; and

generating a user profile with the derived user preferences, the user profile including a keyword mapped to a topic that is associated with a first sub-cluster and that the user did not indicate interest in based upon the keyword mapping to a second topic that is associated with a second sub-cluster and that the user did indicate interest in.

2. The method of claim 1 further comprising:

assigning clusters of questions representing a plurality of topics.

3. The method of claim 2 wherein formulating questions for each cluster comprises: grouping the questions into hierarchical sub-clusters within each cluster, wherein each subsequent level of sub-cluster represents a finer level of granularity in the questions for that sub-cluster.

4. The method of claim 1 wherein no interest is indicated by a negative response or a neutral response from the user.

5. The method of claim 1 further comprising: storing the user profile.

6. The method of claim 5 further comprising: using the user profile to screen new questions for the user.

7. The method of claim 1 further comprising: screening new questions for the user's social connections, using the user profile.

8. An information processing system for deriving a user profile, comprising:

a memory with computer-executable instructions stored therein, the computer-executable instructions causing a computer to perform:

generating a user interface on which a user, for whom no profile exists, interacts in a session;

formulating questions such that a user response to the questions indicates likes and dislikes of the user;

initially presenting the questions representing topics from different high-level categories to the user on the user interface;

receiving responses from the user, and the responses indicating user interest from among the different high-level categories;

iteratively performing the following for each high-level category in which the user expressed an interest:

selecting a next level of cluster of questions from an increasingly narrowed-down subset of clusters, when one or more responses to questions from a previous level of cluster indicated user interest;

presenting questions from the selected next level of cluster;

receiving user responses to the questions; and

deriving user preferences from the user responses;

inhibiting at least one cluster, including one or more sub-clusters within the at least one cluster, from consideration for the session when the user indicates no interest in at least one topic associated with the at least one cluster; and

generating a user profile with the derived user preferences, the user profile including a keyword mapped to a topic that is associated with a first sub-cluster and that the user did not indicate interest in based upon the keyword mapping to a second topic that is associated with a second sub-cluster and that the user did indicate interest in.

9. The information processing system of claim 8 wherein the computer-executable instructions further comprise:

assigning clusters of questions representing a plurality of topics.

10. The information processing system of claim 9 wherein the computer-executable instructions further comprise, for each cluster: grouping the questions into hierarchical sub-clusters within each cluster, wherein each subsequent level of sub-cluster represents a finer level of granularity in the questions for that sub-cluster.

11. The information processing system of claim 8 wherein no interest is indicated by a negative response or a neutral response from the user.

12. The information processing system of claim 8 wherein the computer-executable instructions further comprise: storing the user profile.

13. The information processing system of claim 12 wherein the computer-executable instructions further comprise: using the user profile to screen new questions for the user.

14. The information processing system of claim 12 wherein the computer-executable instructions further comprise: screening new questions for the user's social connections, using the user profile.

15. A computer program product comprising a non-transitory computer-readable storage medium with computer-executable stored therein, the computer-executable instructions comprising:

generating a user interface on which a user interacts in a session;

formulating questions such that a user response to the questions indicates likes and dislikes of the user;

initially presenting the questions representing topics from different high-level categories to the user on the user interface;

receiving responses from the user, the responses indicating user interest from among the different high-level categories;

iteratively performing the following for each high-level category in which the user expressed an interest:

selecting a next level of cluster of questions from an increasingly narrowed-down subset of clusters, when one or more responses to questions from a previous level of cluster indicated user interest;

presenting questions from the selected next level of cluster;

receiving user responses to the questions; and

deriving user preferences from the user responses;

inhibiting at least one cluster, including one or more sub-clusters within the at least one cluster, from consideration for generating of one or more user profiles in association with the session when the user indicates no interest in at least one topic associated with the at least one cluster; and

generating a user profile with the derived user preferences.

16. The computer program product of claim 15 wherein the computer-executable instructions further comprise:

assigning clusters of questions representing a plurality of topics; and

for each cluster, grouping the questions into hierarchical sub-clusters within each cluster, wherein each subsequent level of sub-cluster represents a finer level of granularity in the questions for that sub-cluster.

17. The computer program product of claim 15 wherein the user profile includes a keyword mapped to a topic that is associated with a first sub-cluster and that the user did not indicate interest in based upon the keyword mapping to a second topic that is associated with a second sub-cluster and that the user did indicate interest in.

18. The computer program product of claim 17 wherein the computer-executable instructions further comprise: using the user profile to screen new questions for the user.

Assignments (6)