Methods and systems for context-aware entity correspondence and merging

摘要

A computer-based method for correlating relevant information from multiple entities based on contextual correspondence is described. The method includes receiving, at a computer, information relating to a plurality of the multiple entities, the information including data features and context features, correlating the data features utilizing one or more algorithms running on the computer, correlating the context features utilizing one or more algorithms running on the computer, and aggregating the plurality of the multiple entities based on both a correspondence of the data features and a correspondence of the context features for at least one of storage in a memory associated with the computer and output as data from the computer.

主权项

1. A computer-based method for correlating relevant information from multiple entities based on contextual correspondence, said method comprising:
receiving, at a computer, information relating to a plurality of the multiple entities, the information including data features and context features; correlating the data features utilizing one or more algorithms running on the computer; correlating the context features utilizing the one or more algorithms running on the computer; aggregating the plurality of the multiple entities based on a correspondence of the data features, a correspondence of the context features, and at least one context model for at least one of storage in a memory associated with the computer and output as data from the computer, wherein the at least one context model includes temporal information associated with the context features; receiving a first contextual event including temporal information at a first location associated with the first contextual event; receiving a second contextual event including temporal information at a second location associated with the second contextual event, wherein the first contextual event is different from the second contextual event and the first location is different from the second location; matching a plurality of entities based on the received contextual events; determining a correspondence value for the matched plurality of entities based on a comparison of at least one of the data features or context features in light of the temporal information of the first contextual event and the first location; dynamically updating the correspondence value for the matched plurality of entities based on a comparison of at least one of the data features or context features in light of the temporal information of the second contextual event and the second location; and outputting the matched plurality of entities that exceed a predetermined correspondence value.