Track reconciliation from multiple data sources

• US 10,133,792 B2
• Filed: 06/18/2015
• Issued: 11/20/2018
• Est. Priority Date: 03/31/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Agrawal, Dakshi, Ganti, Raghu K., Jonas, Jeffrey J., Srivatsa, Mudhakar
• Primary Examiner(s): Boccio, Vincent F
• Application Number:
  US14/742,875
• Publication Number:
  US 20150286666A1
• Time in Patent Office: 1,251 Days
• Field of Search: 707746
• CPC Class Codes: G01S 5/00 Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using active systems G01S13/00, G01S15/00, G01S17/00
• CPC Class Codes: G01S 5/0294 Trajectory determination or predictive filtering, e.g. target tracking or Kalman filtering
• CPC Class Codes: G06F 16/2477 Temporal data queries
• CPC Class Codes: G06F 2218/00 Aspects of pattern recognition specially adapted for signal processing
• CPC Class Codes: G06V 20/52 Surveillance or monitoring of activities, e.g. for recognising suspicious objects recognising microscopic objects G06V20/69
• CPC Class Codes: H04W 64/00 Locating users or terminals or network equipment for network management purposes, e.g. mobility management

Abstract

Embodiments relate to reconciling different entity identifiers. A method of reconciling different entity identifiers of a same entity is provided. The method receives a plurality of series of location-time data items from a plurality of tracking systems that each track one or more entities. Each series of location-time data items is associated with an entity identifier. The method categorizes each location-data item into a space-time region. The method generates a track for each of the plurality of series of location-time data items based on the space-time regions into which the location-data items are categorized, and generates a track signature for each of the generated tracks based on a segment of the generated track. The method compares the track signatures to find matching track signatures. Based on a plurality of matching signatures, the method reconciles the plurality of entity identifiers associated with the plurality of matching signatures to a particular entity.

Claims

1. A computer implemented method of reconciling different entity identifiers of a same entity, the method comprising:

receiving, by a processor, a plurality of series of location-time data items from a plurality of tracking systems that each track one or more entities, each series of location-time data items associated with an entity identifier;

categorizing each location-data item into a space-time region;

generating a track for each of the plurality of series of location-time data items based on the space-time regions into which the location-data items are categorized;

generating a track signature for each of the generated tracks based on a segment of the generated track;

comparing the track signatures to find matching track signatures;

based on a plurality of matching signatures, reconciling the plurality of entity identifiers to a same entity; and

responsive to determining a threshold number of tracks, generating a track signature for each of the generated tracks based on a shorter segment of the generated track, wherein the shorter segment is a different size than the segment of the generated track.

2. The method of claim 1 further comprising:

identifying a group of tracks that have a same track signature;

reducing a size of the group using a subsequent track signature generated from the tracks in the group by removing those tracks in the group that do not have the subsequent track signature;

repeating the reducing the size of the group until the size of the group drops down to a threshold number of tracks; and

reconciling the entity identifiers associated with the tracks remaining in the group to a particular entity.

3. The method of claim 2, wherein generating the track signature for each of the generated track based on the shorter segment of the generated track such that more tracks have a same track signature is generated responsive to determining that the size of the group drops below the threshold number of tracks.

4. The method of claim 1, wherein a space-time region includes location information and a time interval, and the location information and the time interval is represented by a location and a time instance, respectively, wherein the categorizing each location-data item comprises quantizing a location and a time instance of a location-time data item into the location and the time instance for the space-time region, respectively.

5. The method of claim 1, wherein a space-time region includes location information and a time interval, wherein the generating a track signature comprises hashing the location information of each space-time region into a hash value using a locality-sensitive hashing technique.