Identifying activities using a hybrid user-activity model
View Patent ↗In a user-activity identification technique, a user's actions are monitored while the user is using a computer. While these user actions are associated with user activities, the user activities are initially unspecified, so the tracked user actions constitute unsupervised data. Then, the tracked user actions are aggregated into subsets (for example, using clustering analysis), and user-activity classifications for the subsets (such as activity labels) are provided by the user, so the subsets constitute supervised data. Subsequently, when additional user actions (which are associated with one or more initially unspecified current user activities) are tracked, they can be associated with one or more of the classified subsets. For example, information about the additional user actions can be mapped in real time (or near real time) to one or more of the subsets using a supervised learning technique. In this way, the one or more current user activities can be identified.
1. A method for identifying one or more current user activities, comprising:
obtaining, by a computer, user-action information associated with a current user activity, wherein the current user activity is initially unspecified, and wherein the user-action information is gathered automatically without user intervention and comprises unsupervised data corresponding to the activity's primitive operations;
identifying, from a plurality of classified user-action information subsets that each comprises supervised data, one or more classified user-action information subsets whose associated user-action information matches the received user-action information;
calculating a weighted sum of the identified matches;
calculating, based on the weighted sum, a probability that a current user activity corresponds to a respective classified user-action information subset;
determining, based on the calculated probability, a classified user-action-information subset which corresponds to the received additional user-action information;
identifying a classified user activity that corresponds to the determined user-action-information subset.
2. The method of claim 1 , wherein obtaining the user-action information involves monitoring user interactions with an electronic device.
3. The method of claim 2 , wherein the obtained user-action information includes one or more of: a keyword associated with a document accessed by the user, the document accessed by the user, an individual with whom the user communicates information, and an appointment in a user's schedule.
4. The method of claim 1 , wherein the user activities include a temporal sequence of user actions that modify content or communicate information.
5. The method of claim 4 , wherein the temporal sequence of user actions occurs over multiple sessions.
6. The method of claim 4 , wherein the user activities are associated with one or more corresponding objectives of a project.
7. The method of claim 1 , wherein determining the association between the at least one labeled user-action-information subset and the additional user-action information involves generating probabilistic scores for the user-action-information subsets based on the additional user-action information.
8. The method of claim 7 , wherein determining the association between the at least one labeled user-action-information subset and the additional user-action information involves selecting a pre-defined group of the user-action-information subsets based on the generated probabilistic scores.
9. The method of claim 7 , wherein the probabilistic scores are generated using corresponding user-action information in the given user-action-information subset that is associated with a time window preceding the one or more current user activities.
10. The method of claim 1 , wherein the obtained user-action information and the received additional user-action information are each associated with tasks;
wherein a given user activity is associated with a temporal sequence of multiple tasks;
wherein the method further comprises identifying transitions between the tasks based on the user-action information or the additional user-action information; and
wherein determining the association between the at least one user-action-information subset and the additional user-action information is based on the identified transitions.
11. The method of claim 1 , wherein determining the association between the at least one labeled user-action-information subset and the additional user-action information is based on contextual information associated with the additional user-action information.
12. The method of claim 1 , wherein the method further comprises updating a time card of the user based on the identified one or more current user activities.
13. The method of claim 1 , wherein the method further comprises providing an activity update to another individual based on the identified one or more current user activities.
14. The method of claim 13 , wherein the activity update alerts the other individual when the user can be contacted without interrupting the one or more current user activities.
15. The method of claim 1 , further comprising:
obtaining user-action information associated with user activities for one or more users, wherein the user activities are initially unspecified and the obtained user-action information comprises unsupervised data; and
clustering the obtained user-action information to generate one or more user-action information subsets.
16. The method of claim 15 , further comprising:
receiving, from a user, an activity label that is to be applied to a respective user-action-information subset; and
classifying the respective user-action-information subset by associating the activity label with the respective user-action-information subset.
17. A non-transitory computer-program product for use in conjunction with a computer system, the computer-program product comprising a computer-readable storage medium and a computer-program mechanism embedded therein to identify one or more current user activities, the computer-program mechanism including:
instructions for obtaining user-action information associated with a current user activity, wherein the current user activity is initially unspecified, and wherein the user-action information is gathered automatically without user intervention and comprises unsupervised data corresponding to the activity's primitive operations;
instructions for identifying, from a plurality of classified user-action information subsets that each comprises supervised data, one or more classified user-action information subsets whose associated user-action information matches the received user-action information;
instructions for calculating a weighted sum of the identified matches;
instructions for calculating, based on the weighted sum, a probability that a current user activity corresponds to a respective classified user-action information subset;
instructions for determining, based on the calculated probability, a classified user-action-information subset which corresponds to the received additional user-action; and
instructions for identifying a classified user activity that corresponds to the determined user-action-information subset.
18. A computer system, comprising:
a processor;
memory; and
a program module, wherein the program module is stored in the memory and configured to be executed by the processor, the program module to identify one or more current user activities, the program module including:
instructions for obtaining user-action information associated with a current user activity, wherein the current user activity is initially unspecified, and wherein the user-action information is gathered automatically without user intervention and comprises unsupervised data corresponding to the activity's primitive operations;
instructions for identifying, from a plurality of classified user-action information subsets that each comprises supervised data, one or more classified user-action information subsets whose associated user-action information matches the received user-action information;
instruction for calculating a weighted sum of the identified matches;
instruction for calculating, based on the weighted sum, a probability that a current user activity corresponds to a respective classified user-action information subset;
instructions for determining, based on the calculated probability, a classified user-action-information subset which corresponds to the received additional user-action information; and
instructions for identifying a classified user activity that corresponds to the determined user-action-information subset.