Method for approximating user task representations by document-usage clustering
View Patent ↗Embodiments of the present invention provide a system for automatically creating a task representation associated with a user task. The system calculates usage footprints of a document based on other applications, documents, and people that have been accessed by the user within a predetermined time frame before and after the user accesses the document. After obtaining usage footprints of a number of documents, the system applies a clustering technique, such as spectral clustering, to create task representations, each including a collection (cluster) of documents and/or applications that are used for accomplishing a particular task. The system also filters the documents based on their average dwell times, and uses user feedback to merge or split different task clusters in order to provide accurate task representations.
1. A computer-executable method for detecting user tasks, the method comprising:
receiving a number of documents;
calculating document-usage footprints for the received documents;
calculating document dwell times for the received documents;
selecting a subset of documents from the received documents such that the document dwell times of the selected documents exceed a predetermined threshold; and
applying a clustering technique to the calculated document-usage footprints of the selected documents to generate user task representations.
2. The method of claim 1 , wherein a document-usage footprint of a respective document is calculated based on access frequencies to other documents within one or more predetermined time periods.
3. The method of claim 2 , wherein the predetermined time periods include at least one of:
a predetermined time period before the respective document is accessed;
a predetermined time period within a time period when the respective document is accessed; and
a predetermined time period after the respective document is accessed.
4. The method of claim 1 , wherein calculating a dwell time of a respective document involves subtracting an idle time from an active time of the document.
5. The method of claim 1 , wherein the clustering technique includes spectral clustering.
6. The method of claim 1 , further comprising:
obtaining user feedback in order to merge and/or split the task representations.
7. A computer-readable storage medium storing instructions that when executed by a computer cause the computer to perform a method for detecting user tasks, the method comprising:
receiving a number of documents;
calculating document-usage footprints for the received documents;
calculating document dwell times for the received documents;
selecting a subset of documents from the received documents such that the document dwell times of the selected documents exceed a predetermined threshold; and
applying a clustering technique to the calculated document-usage footprints of the selected documents to generate user task representations.
8. The computer-readable storage medium of claim 7 , wherein a document-usage footprint of a respective document is calculated based on access frequencies to other documents within one or more predetermined time periods.
9. The computer-readable storage medium of claim 8 , wherein the predetermined time periods include at least one of:
a predetermined time period before the respective document is accessed;
a predetermined time period within a time period when the respective document is accessed; and
a predetermined time period after the respective document is accessed.
10. The computer-readable storage medium of claim 7 , wherein calculating a dwell time of a respective document involves subtracting an idle time from an active time of the document.
11. The computer-readable storage medium of claim 7 , wherein the clustering technique includes spectral clustering.
12. The computer-readable storage medium of claim 7 , wherein the method further comprises obtaining user feedback in order to merge and/or split the task representations.
13. A system for detecting user tasks, the system comprising:
a receiving mechanism configured to receive a number of documents;
a document-usage calculating mechanism configured to calculate document-usage footprints for the received documents;
a dwell-time calculation mechanism configured to calculate document dwell times for the received documents;
a selection mechanism configured to select a subset of documents from the received documents such that the document dwell times of the selected documents exceed a predetermined threshold; and
a clustering mechanism configured to apply a clustering technique to the calculated document-usage footprints of the selected documents to generate user task representations.
14. The system of claim 13 , wherein a document-usage footprint of a respective document is calculated based on access frequencies to other documents within one or more predetermined time periods.
15. The system of claim 14 , wherein the predetermined time periods include at least one of:
a predetermined time period before the respective document is accessed;
a predetermined time period within a time period when the respective document is accessed; and
a predetermined time period after the respective document is accessed.
16. The system of claim 13 , wherein a dwell time of a document is calculated by subtracting an idle time from an active time of the document.
17. The system of claim 13 , wherein the clustering technique includes spectral clustering.
18. The system of claim 13 , further comprising a feedback mechanism configured to obtain user feedback in order to merge and/or split the task representations.