Systems and methods for real-time location-aware recommendations

• US 10,114,901 B2
• Filed: 07/20/2012
• Issued: 10/30/2018
• Est. Priority Date: 07/20/2011
• Status: Active Grant

• Current Assignee: eBay Inc.
• Original Assignee: eBay Inc.
• Inventors: Mitrovic, Ivan
• Primary Examiner(s): Raab, Christopher J
• Application Number:
  US13/554,584
• Publication Number:
  US 20130024471A1
• Time in Patent Office: 2,293 Days
• Field of Search: None
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9537 Spatial or temporal dependent retrieval, e.g. spatiotemporal queries

Abstract

Systems and methods for real-time location-aware recommendations are discussed herein. In an example, a method for generating location-aware recommendations can include receiving a current location of a user, calculating a place graph for the user, traversing the place graph, selecting a recommendation, and communicating the recommendation. In this example, the place graph is calculated based on the current location. Traversing the place graph can identify one or more potential recommendations within a pre-defined distance of the current location.

Claims

1. A method for generating location-aware recommendations, the method comprising:

receiving a search request that includes location data that identifies a current location of a user device;

accessing user profile data for the user, the user profile data including a user interaction history of the user, the user interaction history comprising at least an explicit user input and an implicit user input, the explicit user input including at least a request to promote a place from among a first plurality of places, and the implicit user input including at least an indication that the user visited a web-site associated with the place from among the first plurality of places;

extracting a first set of features from the first plurality of places, the first set of features including attributes of the first plurality of places;

assigning a feature value to each feature among the first set of features, the feature value being based on the implicit user input and the explicit user input of the user, the assigning the feature value to each feature among the first set of features includes applying a first weighting factor to the explicit user input regarding the one or more of the first plurality of places, and a second weighting factor to the implicit input regarding the one or more of the first plurality of places;

generating a feature matrix of the user based on the feature value of each feature among the first set of features;

identifying a second plurality of places within a radius of the current location, the second plurality of places having a second set of features;

generating a place graph of the user, the generating the place graph including projecting the feature matrix of the user onto the second set of features of the second plurality of places;

identifying a place recommendation from among the second plurality of places based on the place graph that includes the feature value from the feature matrix; and

communicating the place recommendation to the user.

2. The method of claim 1, wherein the explicit user input includes ratings of each of the first plurality of places from the user.

3. The method of claim 1, wherein the assigning the feature value to each feature among the first set of features based on explicit user input includes scoring the features based on one or more of the following explicit inputs:

a user supplied rating;

a review;

a related check-in;

or

a user saving a place into an address book.

4. The method of claim 1, wherein the identifying the place recommendation includes filtering the one or more potential place recommendations based on time of day.

5. A system comprising:

one or more processors; and

a memory storing instructions that, when executed by at least one processor among the one or more processors, causes the system to perform operations comprising;

receiving a search request that includes location data that identifies a current location of a user device;

accessing user profile data for the user, the user profile data including a user interaction history of the user, the user interaction history comprising at least an explicit user input and an implicit user input, the explicit user input including at least a request to promote a place from among a first plurality of places, and the implicit user input including at least an indication that the user visited a web-site associated with the place from among the first plurality of places;

extracting a first set of features from the first plurality of places, the first set of features including attributes of the first plurality of places;

assigning a feature value to each feature among the first set of features, the feature value based on the implicit user input and the explicit user input of the user, the assigning the feature value to each feature among the first set of features includes applying a first weighting factor to the explicit user input regarding the one or more of the first plurality of places, and a second weighting factor to the implicit input regarding the one or more of the first plurality of places;

generating a feature matrix of the user based on the feature value of each feature among the first set of features;

identifying a second plurality of places within a radius of the current location, the second plurality of places having a second set of features;

generating a place graph of the user, the generating the place graph including projecting the feature matrix of the user onto the second set of features of the second plurality of places;

identifying a place recommendation from among the second plurality of places based on the place graph that includes the feature value from the feature matrix; and

communicating the place recommendation to the user.

6. The system of claim 5, wherein the explicit user input includes ratings of each of the first plurality of places from the user.

7. A non-transitory machine-readable storage medium containing instructions that, when executed by at least one processor of a machine, cause the machine to perform operations comprising:

receiving a search request that includes location data that identifies a current location of a user device;

accessing user profile data for the user, the user profile data including a user interaction history of the user, the user interaction history comprising at least an explicit user input and an implicit user input, the explicit user input including at least a request to promote a place from among a first plurality of places, and the implicit user input including at least an indication that the user visited a web-site associated with the place from among the first plurality of places;

extracting a first set of features from the first plurality of places, the first set of features including attributes of the first plurality of places;

assigning a feature value to each feature among the first set of features, the feature value based on the implicit user input and the explicit user input of the user, the assigning the feature value to each feature among the first set of features includes applying a first weighting factor to the explicit user input regarding the one or more of the first plurality of places, and a second weighting factor to the implicit input regarding the one or more of the first plurality of places;

generating a feature matrix of the user based on the feature value of each feature among the first set of features;

identifying a second plurality of places within a radius of the current location, the second plurality of places having a second set of features;

generating a place graph of the user, the generating the place graph including projecting the feature matrix of the user onto the second set of features of the second plurality of places;

identifying a place recommendation from among the second plurality of places based on the place graph that includes the feature value from the feature matrix; and

communicating the place recommendation to the user.