Method and system for identifying significant locations through data obtainable from a telecommunication network

• US 9,706,363 B2
• Filed: 04/18/2014
• Issued: 07/11/2017
• Est. Priority Date: 04/18/2014
• Status: Active Grant

• Current Assignee: Telecom Italia S.p.A.
• Original Assignee: Telecom Italia S.p.A.
• Inventors: Galassi, Marco, Mamei, Marco, Colonna, Massimo
• Primary Examiner(s): Sarwar, Babar
• Application Number:
  US15/304,020
• Publication Number:
  US 20170041762A1
• Time in Patent Office: 1,180 Days
• Field of Search: 4554042, 4554121-4142, 455418-4221, 455 411- 412, 4554561, 4554563, 455457, 4555521, 370328, 370338, 370312
• CPC Class Codes: H04L 67/51 Discovery or management thereof, e.g. service location protocol [SLP] or web services
• CPC Class Codes: H04W 4/029 Location-based management or tracking services

Abstract

A method for identifying and locating at least one relevant location visited by at least one individual within a geographical area served by a wireless telecommunication network. The method includes: identifying and clustering events that occurred within a predefined distance from each other and having a similar probability, computing a weight value for each cluster of events identified, comparing the weight value with a threshold weight value, if the weight value is equal to, or greater than, the threshold weight value, identifying the relevant location as belonging to a selected typology of relevant location, and providing an indication of the position of the at least one relevant location based on recorded position data of the events of the cluster, or if the weight value is lower than the threshold weight value, identifying the relevant location as not belonging to the selected typology of relevant location.

Claims

1. A method for identifying and locating at least one relevant location visited by at least one individual within a geographical area served by a wireless telecommunication network, the method comprising:

selecting a predetermined time period over which the identifying and locating of the at least one relevant location have to be performed;

selecting a typology of relevant location to be identified and located;

retrieving recorded time and position data recorded by the telecommunication network and regarding events in which a user equipment carried by the at least one individual interacted with the telecommunication network;

computing a probability that each event for which recorded time and position data have been retrieved occurred in the relevant location of the selected type based on the recorded time data;

identifying and clustering events that occurred within a predefined distance from each other and having a similar probability;

computing a weight value for each identified cluster of events;

comparing the weight value with a threshold weight value,

if the weight value is equal to, or greater than, the threshold weight value, identifying the relevant location as belonging to the selected typology of relevant location, and providing an indication of the position of the at least one relevant location based on the recorded position data of the events of the cluster, or

if the weight value is lower than the threshold weight value, identifying the relevant location as not belonging to the selected typology of relevant location.

2. The method according to claim 1, wherein the retrieving recorded time and position data comprises verifying minimum requirements for the time and position data for ensuring reliability and accuracy of the relevant location identification, wherein the verifying minimum requirements comprises one or more of the following:

verifying if a total number of events within the predetermined time period, associated with the at least one individual is equal to, or greater than, a total events threshold;

verifying if an average number of daily events, within the predetermined time period, associated with the at least one individual is equal to, or greater than, a predetermined average events threshold, and

verifying if a number of events within a predetermined subinterval of the predetermined time period associated with the at least one individual is equal to, or greater than, a subinterval events threshold.

3. The method according to claim 1, wherein the computing a probability comprises exploiting a set of probability functions each of which describes a probability that the at least one individual is at the corresponding relevant location at a time instant defined in the time data during a predetermined subinterval of the predetermined time period.

4. The method according to claim 3, wherein the probability functions are based on statistical analysis data regarding an attendance rate of the selected typology of relevant location.

5. The method according to claim 1, wherein the identifying and clustering further comprises clustering events associated with the at least one individual in a same cluster only if they are sufficiently close in space and time to a centroid of the cluster.

6. The method according to claim 5, wherein the identifying and clustering further comprises exploiting the Hartigan Leader algorithm.

7. The method according to claim 5, wherein the identifying and clustering further comprises, for each event associated with the at least one individual, computing a spatiotemporal distance (std(c_u,e_i)) function expressed as:

std(c_u,e_i)=sd(c_u,e_i)+k×

td(c_u,e_i),

wherein sd(c_u,e_i) is a spatial component of the spatiotemporal distance function, td(c_u,e_i) is a time component of the spatiotemporal distance function, and k is an adjusting parameter for taking into account influence of the time component over the spatial component.

8. The method according to claim 7, wherein the identifying and clustering further comprises computing the spatial component (sd(c_u,e_i)) of the spatiotemporal function as:

sd(c_u,e_i)=max[0,geo(c_u,e_i)−

(r+r_i)],

where geo(c_u,e_i) is the distance between the centroid of the Cluster and the position indicated in the position data regarding the corresponding event, r is an average cluster radius and r_i is a radius comprised in the position data regarding the event.

9. The method according to claim 7, wherein the identifying and clustering further comprises computing the time component (td(c_u,e_i)) of the spatiotemporal function as:

10. The method according to claim 9, wherein the identifying and clustering further comprises comparing a result of the spatiotemporal distance function for the event with a spatiotemporal threshold distance and:

if such result is equal to, or lower than, the spatiotemporal threshold distance, considering the event as belonging to the cluster, whereas

if such result is greater than the spatiotemporal threshold distance, considering the event as not belonging to the cluster.

11. The method according to claim 1, wherein the computing a weight value comprises composing the weight value as a linear combination of a plurality of partial weight values, each partial weight value being designed to take into account statistical aspects that affect the identification and the positioning of the selected relevant location.

12. The method according to claim 11, wherein the partial weight values comprise a first partial weight value (w₁) which takes into account the probability that each event of the cluster has occurred, and is expressed as:

13. The method according to claim 11, wherein the partial weight values comprise a second partial weight value which takes into account the number of predetermined subintervals of the predetermined time period in which events of the cluster occurs.

14. The method according to claim 11, wherein the partial weight values comprise a third partial weight value (w₃) which takes into account a distribution of the occurrence of the events in a portion of the predetermined time period comprising a plurality of predetermined subinterval, and is expressed as:

w₃=T×

H(X),

where H(X) is the Shannon entropy of the cluster in a predetermined subinterval of the predetermined time period, and T is the predetermined time period expressed as a number of predetermined subintervals.

15. The method according to claim 11, further comprising computing a threshold weight value computed as per the weight value but based on an ideal cluster of ideal events associated with an ideal individual which are all generated in the selected relevant location and have a probability greater than zero.

16. The method according to claim 11, further comprising computing a threshold weight value computed as a linear combination of a mean value of the weight values of all clusters identified and a variance thereof.

17. The method according to claim 11, wherein the computing a threshold weight value comprises selecting as a threshold weight value the weight value with the highest value among all the weight values computed.

18. A system coupled to a wireless telecommunication network adapted to implement the method of claim 1, the system comprising:

a computation engine configured to process data retrieved from a mobile telephony network;

a repository configured to store at least one of;

data regarding interactions between the user equipment and the mobile telephony network, computation results generated by the computation engine;

any processing data generated by and/or provided to the system;

an administrator interface configured to modify parameters and/or algorithms used by the computation engine and/or accessing data stored in the repository.

19. The system according to claim 18, further comprising at least one user interface configured to receive inputs from, and to provide output to a user of the system, the user comprising one or more human beings and/or one or more external computing systems subscriber of the services provided by the system.