Methods and systems for scalable group detection from multiple data streams

Abstract

A system, method and computer program product for identifying strong links and discovering hidden relationships among entities, including identifying strong links and discovering hidden relationships among entities, wherein the entities include places, time slots, people, groups, and organizations; and identifying the strong links and discovering the hidden relationships based on low-level data streams, and incomplete and noisy evidence data streams.

What is claimed is: 
     
         1 . A computer implemented system for identifying strong links and discovering hidden relationships among entities, the system comprising:
 the system configured for identifying strong links and discovering hidden relationships among entities, wherein the entities include places, time slots, people, groups, and organizations; and   the system configured for identifying the strong links and discovering the hidden relationships based on low-level data streams, and incomplete and noisy evidence data streams.   
     
     
         2 . The system of  claim 1 , wherein the system is configured to combine knowledge representation and reasoning with approximation and randomization, text mining, machine learning, link discovery and statistical analysis to find a connection among seemingly unrelated entities from data streams. 
     
     
         3 . The system of  claim 1 , wherein the system is configured to identify groups based on knowing a small number of group members by approximation and sampling based on algorithms including Hoeffding bound to reduce a potential error between sampled and non-sampled data to determine if an error is acceptable or not. 
     
     
         4 . The system of  claim 1 , wherein the system is configured to expand a hybrid link discovery model from a static database to a multi-stream database. 
     
     
         5 . A method for a computer implemented system for identifying strong links and discovering hidden relationships among entities, the method comprising:
 identifying with the system strong links and discovering hidden relationships among entities, wherein the entities include places, time slots, people, groups, and organizations; and   identifying the strong links and discovering the hidden relationships with the system based on low-level data streams, and incomplete and noisy evidence data streams.   
     
     
         6 . The method of  claim 5 , further comprising combining with the system knowledge representation and reasoning with approximation and randomization, text mining, machine learning, link discovery and statistical analysis to find a connection among seemingly unrelated entities from data streams. 
     
     
         7 . The method of  claim 5 , further comprising identifying with the system groups based on knowing a small number of group members by approximation and sampling based on algorithms including Hoeffding bound to reduce a potential error between sampled and non-sampled data to determine if an error is acceptable or not. 
     
     
         8 . The method of  claim 5 , further comprising expanding with the system a hybrid link discovery model from a static database to a multi-stream database. 
     
     
         9 . A computer program product for a computer implemented system for identifying strong links and discovering hidden relationships among entities, and including one or more computer readable instructions embedded on a non-transitory, tangible computer readable medium and configured to cause one or more computer processors to perform the steps of:
 identifying with the system strong links and discovering hidden relationships among entities, wherein the entities include places, time slots, people, groups, and organizations; and   identifying the strong links and discovering the hidden relationships with the system based on low-level data streams, and incomplete and noisy evidence data streams.   
     
     
         10 . The computer program product of  claim 9 , further comprising combining with the system knowledge representation and reasoning with approximation and randomization, text mining, machine learning, link discovery and statistical analysis to find a connection among seemingly unrelated entities from data streams. 
     
     
         11 . The computer program product of  claim 9 , further comprising identifying with the system groups based on knowing a small number of group members by approximation and sampling based on algorithms including Hoeffding bound to reduce a potential error between sampled and non-sampled data to determine if an error is acceptable or not. 
     
     
         12 . The computer program product of  claim 9 , further comprising expanding with the system a hybrid link discovery model from a static database to a multi-stream database.