Keyword refinement in temporally evolving online media

• US 9,910,921 B2
• Filed: 02/28/2013
• Issued: 03/06/2018
• Est. Priority Date: 02/28/2013
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Krishnapuram, Raguhuram, Mehta, Sameep, Shyamsundar, Chawda Bhupesh, Subramaniam, L Venkata
• Primary Examiner(s): Mofiz, Apu
• Assistant Examiner(s): Agharahimi, Farhad
• Application Number:
  US13/780,874
• Publication Number:
  US 20140244611A1
• Time in Patent Office: 1,832 Days
• Field of Search: None
• CPC Class Codes: G06F 16/3322 using system suggestions G06F16/3325 takes precedence
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

Methods and arrangements for keyword refinement and enhancement. There is received an initial keyword list comprising one or more keywords. Information is harvested from one or more information feeds, and an item is ascertained from the harvested information. One or more keywords from the initial keyword list are associated with the item. One or more new keywords are developed based on the associating of one or more keywords from the initial keyword list with the item. Other variants and embodiments are broadly contemplated herein.

Claims

1. A method of keyword enhancement, said method comprising:

utilizing a processor to execute computer code configured to perform the steps of;

receiving an initial keyword list comprising one or more keywords;

harvesting, using the initial keyword list, information from one or more social media information feeds and ascertaining a plurality of events from the harvested information, the events having a location and time;

clustering, using at least one of the one or more keywords as a cluster center, the events based on location and time, each cluster being represented by a feature vector, wherein each feature vector comprises a set of keywords, the location, and the time associated with the cluster;

identifying a temporal evolution state of the event and updating the feature vector associated with the cluster with the temporal evolution state; and

updating the set of keywords associated with each cluster, wherein the updating the set of keywords comprises removing one or more keywords for event clusters identified as declining based upon the temporal evolution state of the event and adding one or more keywords for event clusters identified as growing based upon the temporal evolution state of the event;

the updating the set of keywords comprising accepting the updated feature vector associated with the cluster, generating a candidate list, and comparing information from one or more social media information feeds with the candidate list to generate new keywords for growing event clusters.

2. The method according to claim 1, comprising determining a score for each of the events, the score indicating a strength of association between the information from one or more social media information feeds and the event.

3. The method according to claim 1, wherein said updating comprises applying a score to one or more possible new keywords and applying a threshold test to the applied score.

4. The method according to claim 3, wherein the score of the one or more possible new keywords is based on the score relative to the feature vector.

5. The method according to claim 1, wherein said updating comprises applying a score to one or more possible new keywords and applying a threshold test to the applied score.

6. The method according to claim 1, comprising updating the initial keyword list with the one or more new keywords.

7. The method according to claim 1, wherein the temporal evolution state comprises one member selected from the group consisting of:

growing, declining, and no change.

8. The method according to claim 7, wherein said developing comprises selecting one or more new keywords with a bias towards a temporal evolution state of growing.

9. An apparatus for keyword enhancement, said apparatus comprising:

at least one processor; and

a computer readable storage medium having computer readable program code embodied therewith and executable by the at least one processor, the computer readable program code comprising;

computer readable program code configured to receive an initial keyword list comprising one or more keywords;

computer readable program code configured to harvest, using the initial keyword list, information from one or more social media information feeds and ascertain a plurality of events from the harvested information, the events having a location and time;

computer readable program code configured to cluster, using at least one of the one or more keywords as a cluster center, the events based on location and time, each cluster being represented by a feature vector, wherein each feature vector comprises a set of keywords, the location, and the time associated with the cluster;

computer readable program code configured to identify a temporal evolution state of the event and updating the feature vector associated with the cluster with the temporal evolution state; and

computer readable program code configured to update the set of keywords associated with each cluster, wherein the updating the set of keywords comprises removing one or more keywords for event clusters identified as declining based upon the temporal evolution state of the event and adding one or more keywords for event clusters identified as growing based upon the temporal evolution state of the event;

the updating the set of keywords comprising accepting the updated feature vector associated with the cluster, generating a candidate list, and comparing information from one or more social media information feeds with the candidate list to generate new keywords for growing event clusters.

10. A computer program product for keyword enhancement, said computer program product comprising:

a non-transitory computer readable storage medium having computer readable program code embodied therewith, the computer readable program code comprising;

computer readable program code configured to receive an initial keyword list comprising one or more keywords;

computer readable program code configured to harvest, using the initial keyword list, information from one or more social media information feeds and ascertain a plurality of events from the harvested information, the events having a location and time;

computer readable program code configured to cluster, using at least one of the one or more keywords as a cluster center, the events based on location and time, each cluster being represented by a feature vector, wherein each feature vector comprises a set of keywords, the location, and the time associated with the cluster;

computer readable program code configured to identify a temporal evolution state of the event and updating the feature vector associated with the cluster with the temporal evolution state; and

computer readable program code configured to update the set of keywords associated with each cluster, wherein the updating the set of keywords comprises removing one or more keywords for event clusters identified as declining based upon the temporal evolution state of the event and adding one or more keywords for event clusters identified as growing based upon the temporal evolution state of the event;

the updating the set of keywords comprising accepting the updated feature vector associated with the cluster, generating a candidate list, and comparing information from one or more social media information feeds with the candidate list to generate new keywords for growing event clusters.

11. The computer program product according to claim 10, wherein said computer readable program code is configured to determine a score for each of the events, the score indicating a strength of association between the information from one or more social media information feeds and the event.

12. The computer program product according to claim 11, wherein said updating comprises applying a score to one or more possible new keywords and applying a threshold test to the applied score.

13. The computer program product according to claim 12, wherein the score of the one or more possible new keywords is based on the score relative to the feature vector.

14. The computer program product according to claim 10, wherein said updating comprises applying a score to one or more possible new keywords and applying a threshold test to the applied score.

15. The computer program product according to claim 10, wherein said computer readable program code is configured to update the initial keyword list with the one or more new keywords.

16. A method comprising:

obtaining information relative to one or more events from a plurality of external social media information feeds;

comparing the obtained information to an input keyword list;

determining one or more events from the obtained information and matching the one or more events to one or more keywords from the input keyword list;

quantifying a correlation between the one or more events and the one or more matched keyword;

identifying a temporal evolution state of the one or more events; and

based on the quantified correlation, adding one or more new keywords to the one or more events identified as growing based upon the temporal evolution state of the one or more events and classifying the new keywords with a temporal evolution state, wherein adding one or more new keywords comprises accepting the one or more keywords matched with the one or more events, generating a candidate list of keywords for each of the one or more events, and comparing information from one or more social media information feeds with the candidate list to generate new keywords for the one or more events; and

based on the quantified correlation, removing one or more keywords from the one or more events identified as declining based upon the temporal evolution state of the one or more events.