METHOD FOR AUTOMATICALLY INFERRING PLACE PROPERTIES BASED ON SPATIAL ACTIVITY DATA USING BAYESIAN MODELS
A method for automatically inferring place properties is provided. The method includes (i) obtaining one or more event data streams associated with one or more entities, (ii) identifying one or more locations associated with each of the one or more event data streams, (iii) deriving at least one contextual event based on a spatial activity of each of the one or more entities and attributes of at least one entity visiting the one or more locations or a weather data, (iv) inferring a land-use and at least one place property of the one or more locations based on the one or more event data streams using at least one data driven model and the at least one contextual event, and (v) re-estimating the at least one place property of the one or more locations when different value for the at least one of place property is inferred.
1 . A processor-implemented method for automatically inferring one or more place properties based on partially observable entity spatial activity data using one or more data driven models, said method comprising:
obtaining, in real-time, a plurality of event data streams associated with a plurality of entities from one or more independently controlled data sources at different spatio-temporal resolutions, wherein the plurality of event data streams comprise at least one of a timestamped data, a spatial data or at least one entity identifier, wherein the plurality of event data streams partially characterize spatial activity of each of the plurality of entities temporally;
identifying one or more locations associated with each of the plurality of event data streams based on at least one of an entity context, a location context or a global context derived from the plurality of event data streams;
deriving at least one contextual event that affects at least one place property of the one or more locations based on at least one of (i) the spatial activity of each of the plurality of entities temporally associated with the one or more locations, (ii) attributes of at least one entity visiting the one or more locations using a global geo-spatial model, or (iii) a weather data;
automatically inferring a land-use and the at least one place property of the one or more locations based on the plurality of event data streams using at least one data driven model of a place property and the at least one contextual event; and
re-estimating the at least one place property of the one or more locations when the plurality of event data streams infer a different value for the at least one of place property of the one or more locations using the at least one data driven model.
2 . The method of claim 1 , wherein the at least one place property of the one or more locations comprises a plurality of static place properties and a plurality of dynamic place properties, wherein the plurality of static place properties comprise at least one of an address, a latitude, a longitude, an altitude, a building polygon, or a mailbox number, wherein the plurality of dynamic place properties comprise at least one of (i) a frequency of human traffic, (ii) proximity of starting points of visitors, (iii) a location of the starting points of the visitors, (iv) a nature of the starting points of the visitors, (v) a stay time at a different time of a day, (vi) previous top locations and next top locations, (vii) other similar places, or (viii) open spaces.
3 . The method of claim 1 , wherein the method further comprises consolidating a nature of the one or more locations by aggregating or disaggregating the one or more locations with respect to one or more spatial scales, wherein the one or more locations vary from a point to a polygon.
4 . The method of claim 1 , wherein the method further comprises running the at least one data driven model of the place property on a near daily basis and updating at least one metric of the at least one place property of the one or more locations for different time scales, wherein the different time scales comprise a day, a week, a month, a quarter, or a year.
5 . The method of claim 1 , wherein the at least one contextual event comprises at least one of a transportation dynamic, a weather dynamic, or a community dynamic.
6 . The method of claim 1 , wherein the method further comprises inferring the land-use and the at least one place property in response to incoming queries with locations and spatial regions that are not in the identified one or more locations.
7 . The method of claim 1 , wherein the at least one data driven model of the place property comprises at least one of (i) a competitive analysis model, (ii) a visitor count model, (iii) a dwell time determination model, (iv) a catchment area determination model, (v) a similar places analysis model, or (vi) a place pre-post visit attribution model.
8 . The method of claim 1 , wherein the method further comprises (i) extrapolating the spatial data with census data or (ii) interpolating the spatial data with the census data before downstream processing.
9 . The method of claim 1 , wherein the method further comprises filtering and standardizing the plurality of event data streams into a common representational format before downstream processing.
10 . The method of claim 1 , wherein the at least one data driven model uses power law effects while inferring the at least one place property of the one or more locations.
11 . A system for automatically inferring one or more place properties based on partially observable entity spatial activity data using one or more data driven models, said system comprising:
a processor; and
a memory that stores a set of instructions, which when executed by the processor, causes to perform:
obtaining, in real-time, a plurality of event data streams associated with a plurality of entities from one or more independently controlled data sources at different spatio-temporal resolutions, wherein the plurality of event data streams comprise at least one of a timestamped data, a spatial data or at least one entity identifier, wherein the plurality of event data streams partially characterize spatial activity of each of the plurality of entities temporally;
identifying one or more locations associated with each of the plurality of event data streams based on at least one of an entity context, a location context or a global context derived from the plurality of event data streams;
deriving at least one contextual event that affects at least one place property of the one or more locations based on at least one of (i) the spatial activity of each of the plurality of entities temporally associated with the one or more locations, (ii) attributes of at least one entity visiting the one or more locations using a global geo-spatial model, or (iii) a weather data;
automatically inferring a land-use and the at least one place property of the one or more locations based on the plurality of event data streams using at least one data driven model of a place property and the at least one contextual event; and
re-estimating the at least one place property of the one or more locations when the plurality of event data streams infer a different value for the at least one of place property of the one or more locations using the at least one data driven model.
10 . The system of claim 11 , wherein the at least one place property of the one or more locations comprises a plurality of static place properties and a plurality of dynamic place properties, wherein the plurality of static place properties comprises at least one of an address, a latitude, a longitude, an altitude, a building polygon, or a mail box number, wherein the plurality of dynamic place properties comprises at least one of (i) a frequency of human traffic, (ii) proximity of starting points of visitors, (iii) a location of the starting points of the visitors (iv) a nature of the starting points of the visitors (v) a stay time at a different time of a day, (vi) previous top locations and next top locations (vii) other similar places, or (viii) open spaces.
13 . The system of claim 11 , wherein the processor is further configured to consolidate a nature of the one or more locations by aggregating or disaggregating the one or more locations with respect to one or more spatial scales, wherein the one or more locations vary from a point to a polygon.
14 . The system of claim 11 , wherein the processor is further configured to run the at least one data driven model of the place property on a near daily basis and updating at least one metric of the at least one place property of the one or more locations for different time scales, wherein the different time scales comprise a day, a week, a month, a quarter, or a year.
15 . The system of claim 11 , wherein the at least one contextual event comprises at least one of a transportation dynamic, a weather dynamic or a community dynamic.
16 . The system of claim 11 , wherein the processor is further configured to infer the land-use and the at least one place property in response to incoming queries with locations and spatial regions that are not in the identified one or more locations.
17 . The system of claim 11 , wherein the at least one data driven model of the place property comprises at least one of (i) a competitive analysis model, (ii) a visitor count model, (iii) a dwell time determination model, (iv) a catchment area determination model, (v) a similar places analysis model, or (vi) a place pre-post visit attribution model.
18 . The system of claim 11 , wherein the processor is further configured to (i) extrapolate the spatial data with census data or (ii) interpolate the spatial data with the census data before downstream processing.
19 . The system of claim 11 , wherein the processor is further configured to filter and standardize the plurality of event data streams into a common representational format before downstream processing.
20 . A one or more non-transitory computer-readable storage mediums storing the one or more sequences of instructions, which when executed by the one or more processors, causes to perform a method for automatically inferring one or more place properties based partially observable on entity spatial activity data using one or more data driven models, said method comprising:
obtaining, in real-time, a plurality of event data streams associated with a plurality of entities from one or more independently controlled data sources at different spatio-temporal resolutions, wherein the plurality of event data streams comprise at least one of a time stamped data, a spatial data or at least one entity identifier, wherein the plurality of event data streams partially characterize spatial activity of each of the plurality of entities temporally;
identifying one or more locations associated with each of the plurality of event data streams based on at least one of an entity context, a location context or a global context derived from the plurality of event data streams;
deriving at least one contextual event that affects at least one place property of the one or more locations based on at least one of (i) the spatial activity of each of the plurality of entities temporally associated with the one or more locations, (ii) attributes of at least one entity visiting the one or more locations using a global geo-spatial model, or (iii) a weather data;
automatically inferring a land-use and the at least one place property of the one or more locations based on the plurality of event data streams using at least one data driven model of a place property and the at least one contextual event; and
re-estimating the at least one place property of the one or more locations when the plurality of event data streams infer a different value for the at least one of place property of the one or more locations using the at least one data driven model.