Event detection through text analysis using trained event template models

Patent metadata

Field	Value
Publication number	US-2016019470-A1
Application number	US-201514867138-A
Country	US
Kind code	A1
Filing date	Sep 28, 2015
Priority date	Dec 2, 2013
Publication date	Jan 21, 2016
Grant date	—

Abstract

A system and method for detecting events based on input data from a plurality of sources. The system may receive input from a plurality of sources containing information about possible events. A method for event detection involves pre-processing and normalizing a data input from a plurality of sources, extracting and disambiguating events and entities, associate event and entities, correlate events and entities associated from a data input to results from a different data sources to determine if an event has occurred, and store the detected events in a data storage.

First claim

What is claimed is:

1 . A method comprising:
extracting, by a server, a first entity from a first data of a first data source and a second entity from a second data of a second data source; identifying, by the server, a first event model candidate from the first data and a second event model candidate from the second data; comparing, by the server, the first event model candidate against a set of event models stored in a first database and the second event model candidate against the set of event models; determining, by the server, a first score indicative of a first probability of a first match for the first event model based on the comparing and a second score indicative of a second probability of a second match for the second event model based on the comparing; associating, by the server, the first score, the first entity, and the first event model candidate as a first spatial-temporal grouping and the second score, the second entity, and the second event model candidate as a second spatial-temporal grouping; validating, by the server, an event based on an overlap of at least one of the first entity or the first event model candidate from the first spatial-temporal grouping and at least one of the second entity or the second event model candidate from the second spatial-temporal grouping and based on the first score from the first spatial-temporal grouping and the second score from the first spatial-temporal grouping exceeding a threshold representative of the event actually occurring; and storing, by the server, a record associated with the event in a second database based on the validating.

2 . The method of claim 1 , wherein the first database and the second database are one database.

3 . The method of claim 1 , wherein the first database and the second database are different databases.

4 . The method of claim 1 , wherein at least one of the first database or the second database is an in-memory database.

5 . The method of claim 1 , further comprising:
pushing, by the server, a message to a client responsive to the storing, wherein the message is informative of the record.

6 . The method of claim 5 , wherein the pushing is based on a rule set by the client.

7 . The method of claim 1 , wherein at least one of the first data source or the second data source comprises at least one of social media network, a newsfeed, a blog hosting server, an education portal, or a document.

8 . The method of claim 1 , further comprising:
pre-processing, by the server, at least one of the first data or the second data at least one of before the extracting the first entity and the second entity or before the extracting the first event model candidate and the second event model candidate.

9 . The method of claim 1 , wherein at least one model from the set of event models is based on an information manually provided to the at least one model by a user.

10 . The method of claim 1 , wherein the extracting the first entity and the second entity comprises disambiguating at least one of the first entity or the second entity.

11 . A system comprising:
a server; a first database in communication with the server; a second database in communication with the server, wherein the server is configured to:
extract a first entity from a first data of a first data source and a second entity from a second data of a second data source,
identify a first event model candidate from the first data and a second event model candidate from the second data,
compare the first event model candidate against a set of event models stored in the first database and the second event model candidate against the set of event models,
determine a first score indicative of a first probability of a first match for the first event model based on the comparison and a second score indicative of a second probability of a second match for the second event model based on the comparison,
associate the first score, the first entity, and the first event model candidate as a first spatial-temporal grouping and the second score, the second entity, and the second event model candidate as a second spatial-temporal grouping,
validate an event based on an overlap of at least one of the first entity or the first event model candidate from the first spatial-temporal grouping and at least one of the second entity or the second event model candidate from the second spatial-temporal grouping and based on the first score from the first spatial-temporal grouping and the second score from the first spatial-temporal grouping exceeding a threshold representative of the event actually occurring, and
store a record associated with the event in the second database based on the validating.

12 . The system of claim 11 , wherein the first database and the second database are one database.

13 . The system of claim 11 , wherein the first database and the second database are different databases.

14 . The system of claim 11 , wherein at least one of the first database or the second database is an in-memory database.

15 . The system of claim 11 , wherein the server is configured to push a message to a client responsive to the storage, wherein the message is informative of the record.

16 . The system of claim 15 , wherein the pushing is based on a rule set by the client.

17 . The system of claim 11 , wherein at least one of the first data source or the second data source comprises at least one of social media network, a newsfeed, a blog hosting server, an education portal, or a document.

18 . The system of claim 11 , wherein the server is configured to pre-process at least one of the first data or the second data at least one of before the extraction of the first entity and the second entity or before the extraction of the first event model candidate and the second event model candidate.

19 . The system of claim 11 , wherein at least one model from the set of event models is based on an information manually provided to the at least one model by a user.

20 . The system of claim 11 , wherein the extraction of the first entity and the second entity comprises the server disambiguating at least one of the first entity or the second entity.

Assignees

• Qbase Llc

Inventors

• Lightner Scott
• Dave Rakesh
• Boddhu Sanjay
• Flagg Robert

Classifications

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06F40/40

  Processing or translation of natural language (natural language analysis G06F40/20; semantic analysis G06F40/30) · CPC title

• H04L67/10

  in which an application is distributed across nodes in the network (software deployment G06F8/60; multiprogramming arrangements G06F9/46) · CPC title

• G06F16/35

  Clustering; Classification · CPC title

• G06F40/284

  Lexical analysis, e.g. tokenisation or collocates · CPC title