Processing a content item with regard to an event and a location

• US 9,201,880 B2
• Filed: 12/10/2012
• Issued: 12/01/2015
• Est. Priority Date: 06/29/2007
• Status: Active Grant

• Current Assignee: Pulsepoint Incorporated (Pulsepoint Holdings LLC)
• Original Assignee: Allvoices, Inc. (Pulsepoint Holdings LLC)
• Inventors: Tareen, Amra Q., Sood, Sanjay C., Birnbaum, Lawrence A., Hammond, Kristian J., Sundelof, Erik
• Primary Examiner(s): VU, BAI DUC
• Application Number:
  US13/709,711
• Publication Number:
  US 20130246440A1
• Time in Patent Office: 1,086 Days
• Field of Search: 707/999.005, 707/999.102, 707/711
• US Class Current: 1/1
• CPC Class Codes: G06F 16/219 Managing data history or versioning querying versioned data G06F16/2474; querying temporal data G06F16/2477
• CPC Class Codes: G06F 16/38 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/40 of multimedia data, e.g. slideshows comprising image and additional audio data retrieval of still image data G06F16/50; retrieval of audio data G06F16/60; retrieval of video data G06F16/70
• CPC Class Codes: G06F 16/489 using time information
• CPC Class Codes: G06F 16/9537 Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
• CPC Class Codes: G06Q 10/00 Administration; Management

Abstract

Associating a content item with an event is disclosed. A location associated with a received content item is determined. The received content item is associated with an event, at least in part based on an indicia of relatedness, other than the determined location, between the received content item and the event. A criterion that the indicia of relatedness is required to satisfy for the content item to be determined to be associated with the event has a lower value if the determined location associated with the received content item has a first degree of correspondence to a location associated with the event than if the determined location associated with the received content item has a second, lower degree of correspondence to the location associated with the event.

Claims

1. A method for associating a content item with an event, comprising:

determining, by a computer having a processor and non-transitory computer memory including instructions executable by the processor, a location associated with a received content item from a source communicatively connected to the computer;

selecting a plurality of previously processed content items, each associated with a previously identified event;

for each of the selected plurality of previously processed content items;

determining a location overlap between the location associated with the received content item and a location associated with the previously processed content item;

determining a matching threshold based on the location overlap;

determining a score representing a degree of non-location based similarity between the received content item and the previously processed content item; and

retaining the event associated with the previously processed content item when the score is greater than the matching threshold;

determining an overall score for each retained event of one or more retained events based at least on an average of the determined scores of the previously processed content items associated with the respective retained event;

selecting an event from the one or more retained events based on the overall scores of the one or more retained events; and

associating, by the computer, the received content item with the selected event.

2. The method as in claim 1, wherein the degree of non-location based similarity between the received content item and the previously processed content item relates to a frequency-based overlap.

3. The method as in claim 2, wherein the frequency-based overlap is determined based on candidate bi-grams and target bi-grams.

4. The method as in claim 1, wherein the degree of non-location based similarity between the received content item and the previously processed content item relates to a unigram similarity.

5. The method as in claim 4, wherein the unigram similarity comprises a similarity between unigrams associated with the received content item and unigrams associated with a content item associated with the event.

6. The method as in claim 1, wherein the degree of non-location based similarity between the received content item and the previously processed content item relates to an entity overlap.

7. The method as in claim 6, wherein the entity overlap comprises a frequency based overlap between named entities associated with the received content item and named entities associated with the previously processed content item.

8. A system for associating a content item with an event, comprising:

a location determiner running on one or more computers each having a processor and non-transitory computer memory including instructions executable by the processor for determining a location associated with a received content item; and

a content item associater running on the one or more computers, the non-transitory computer memory including further instructions executable by the processor for associating the received content item with an event, wherein the further instructions comprise;

selecting a plurality of previously processed content items, each associated with a previously identified event;

for each of the selected plurality of previously processed content items;

determining a location overlap between the location associated with the received content item and a location associated with the previously processed content item;

determining a matching threshold based on the location overlap;

determining a score representing a degree of non-location based similarity between the received content item and the previously processed content item; and

retaining the event associated with the previously processed content item when the score is greater than the matching threshold;

determining an overall score for each retained event of one or more retained events based at least on an average of the determined scores of the previously processed content items associated with the respective retained event;

selecting an event from the one or more retained events based on the overall scores of the one or more retained events; and

associating, by the one or more computers, the received content item with the selected event.

9. The system of claim 8, wherein the degree of non-location based similarity between the received content item and the previously processed content item relates to a frequency-based overlap.

10. The system of claim 9, wherein the frequency-based overlap is determined based on candidate bi-grams and target bi-grams.

11. The system of claim 8, wherein the degree of non-location based similarity between the received content item and the previously processed content item relates to a unigram similarity.

12. The system of claim 11, wherein the unigram similarity comprises a similarity between unigrams associated with the received content item and unigrams associated with a content item associated with the event.

13. The system of claim 8, wherein the degree of non-location based similarity between the received content item and the previously processed content item relates to an entity overlap.

14. The system of claim 13, wherein the entity overlap comprises a frequency based overlap between named entities associated with the received content item and named entities associated with the previously processed content item.

15. A computer program product for associating a content item with an event, the computer program product embodied in a non-transitory computer readable storage medium comprising computer instructions executable by one or more computers for:

determining a location associated with a received content item;

selecting a plurality of previously processed content items, each associated with a previously identified event;

for each of the selected plurality of previously processed content items;

determining a location overlap between the determined location associated with the received content item and a location associated with the previously processed content item;

determining a matching threshold based on the location overlap;

determining a score representing a degree of non-location based similarity between the received content item and the previously processed content item; and

retaining the event associated with the previously processed content item when the score is greater than the matching threshold;

determining an overall score for each retained event of one or more retained events based at least on an average of the determined scores of the previously processed content items associated with the respective retained event;

associating the received content item with an event, by selecting an event from the one or more retained events based on the overall scores of the one or more retained events.

16. The computer program product of claim 15, wherein the degree of non-location based similarity between the received content item and the previously processed content item relates to a frequency-based overlap.

17. The computer program product of claim 16, wherein the frequency-based overlap is determined based on candidate bi-grams and target bi-grams.

18. The computer program product of claim 15, wherein the degree of non-location based similarity between the received content item and the previously processed content item relates to a unigram similarity.

19. The computer program product of claim 18, wherein the unigram similarity comprises a similarity between unigrams associated with the received content item and unigrams associated with a content item associated with the event.

20. The computer program product of claim 15, wherein the degree of non-location based similarity between the received content item and the previously processed content item relates to an entity overlap.

21. The computer program product of claim 20, wherein the entity overlap comprises a frequency based overlap between named entities associated with the received content item and named entities associated with the previously processed content item.