Event detection

• US 8,463,789 B1
• Filed: 08/27/2010
• Issued: 06/11/2013
• Est. Priority Date: 03/23/2010
• Status: Active Grant

• Current Assignee: Aurea Software Incorporated (ESW Capital, LLC)
• Original Assignee: FirstRain, Inc. (Ignite Technologies Incorporated)
• Inventors: Betz, Martin, Rao, Prashant, Joshi, Ashutosh, Gupta, Aparna, Mohanty, Binay, Upadhyay, Jalvin, Arora, Rajiv, Prospero, Michael, Cooke, David
• Primary Examiner(s): Kerzhner, Aleksandr
• Application Number:
  US12/870,689
• Time in Patent Office: 1,019 Days
• Field of Search: 707/738, 707/999.002
• US Class Current: 707/738
• CPC Class Codes: G06F 16/93 Document management systems
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 2218/12 Classification; Matching
• CPC Class Codes: G06F 3/0482 Interaction with lists of selectable items, e.g. menus
• CPC Class Codes: G06V 30/40 Document-oriented image-based pattern recognition

Abstract

Some embodiments provide a method for identifying an event for a particular category. The method classifies several documents as relevant to several different categories. The method identifies a number of documents relevant to the particular category for a current time period and a background time period. Based on a comparison of the number of documents from the current time period relevant to the particular category and the number of documents from the background time period relevant to the particular category, the method identifies an event for the category for the current time period. Some embodiments calculated a score for the event, and normalize the score based on an average number of documents relevant to each of a set of related categories including the particular category.

Claims

1. A machine readable medium storing a program which when executed by at least one processing unit of a research system identifies an event for a category, the program comprising sets of instructions for:

for each of a plurality of different pre-identified categories for which the research system stores retrievable data, classifying a first set of documents from a current time period and a second set of documents from a background time period as relevant to the category, wherein the current time period and the background time period are separated by a buffer time period in order to isolate the current time period from the background time period;

for each of the plurality of categories, calculating a score for the category for the current time period that quantifies a relative difference in a size of the first set of documents from the current time period and a size of the second set of documents from the background time period;

when the calculated score for a particular category is above a threshold, determining the occurrence of an event for the category in the current time period; and

storing data in the research system indicating the occurrence of an event for each of the plurality of categories for which the calculated score is above the threshold.

2. The machine readable medium of claim 1, wherein the set of instructions for calculating the score for a particular category comprises sets of instructions for:

identifying a size of the first set of documents and a size of the second set of documents; and

comparing the size of the first set of documents with the size of the second set of documents.

3. The machine readable medium of claim 2, wherein the current time period is a current day, wherein the set of instructions for comparing comprises sets of instructions for:

averaging the size of the second set of documents per day of the background time period;

subtracting the average from the size of the first set of documents to calculate a difference; and

dividing the calculated difference by a standard deviation of the documents per day of the background time period.

4. The machine readable medium of claim 1, wherein the current time period is a particular day of the week, wherein the background time period is a set of previous instances of the same particular day of the week as the current time period.

5. The machine readable medium of claim 2, wherein the program further comprises a set of instructions for classifying the event into one of a plurality of event types, wherein the plurality of event types comprises non-events, low events, medium events, and high events.

6. The machine readable medium of claim 1, wherein the set of instructions for classifying a particular document as relevant to a particular category comprises sets of instructions for:

calculating a relevancy score for the document based on an analysis of content of the document; and

determining whether the calculated relevancy score is greater than a minimum relevancy threshold for the category.

7. The machine readable medium of claim 1, wherein the program further comprises sets of instructions for determining, for each of the documents classified as relevant to a particular category, whether the document is a duplicate of another of the documents classified as relevant to the particular category.

8. The machine readable medium of claim 7, wherein the pre-identified categories comprise one or more of companies, industries, people, and products.

9. A machine-implemented method for identifying an event for a category, the method comprising:

for each of a plurality of different pre-identified categories for which a research system stores retrievable data, classifying a first set of documents from a current time period and a second set of documents from a background time period as relevant to the category, wherein the current time period and the background time period are separated by a buffer time period in order to isolate the current time period from the background time period;

for each of the plurality of categories, calculating a score for the category for the current time period that quantifies a relative difference in a size of the first set of documents from the current time period and a size of the second set of documents from the background time period;

when the calculated score for a particular category is above a threshold, determining the occurrence of an event for the category in the current time period; and

storing data in the research system indicating the occurrence of an event for each of the plurality of categories for which the calculated score is above the threshold.

10. The method of claim 9, wherein calculating the score comprises:

identifying a size of the first set of documents and a size of the second set of documents; and

comparing the size of the first set of documents with the size of the second set of documents.

11. The method of claim 10, wherein the current time period is a current day, wherein comparing comprises:

averaging the size of the second set of documents per day of the background time period;

subtracting the average from the size of the first set of documents to calculate a difference; and

dividing the calculated difference by a standard deviation of the documents per day of the background time period.

12. The method of claim 9, wherein the current time period is a particular day of the week, wherein the background time period is a set of previous instances of the same particular day of the week as the current time period.

13. The method of claim 9 further comprising classifying the event into one of a plurality of event types.

14. The method of claim 13, wherein the plurality of event types comprises non-events, low events, medium events, and high events.

15. The method of claim 9, wherein classifying a particular document as relevant to a particular category comprises:

calculating a relevancy score for the document based on an analysis of content of the document; and

determining whether the calculated relevancy score is greater than a minimum relevancy threshold for the category.

16. The method of claim 9 further comprising, for each of the documents classified as relevant to a particular category, determining whether the document is a duplicate of another of the documents classified as relevant to the particular category.

17. The method of claim 16, wherein the pre-identified categories comprise one or more of companies, industries, people, and products.