Context-based queryless presentation of recommendations
Techniques for context-based queryless presentation of items are described. Without receiving an explicit query, an inferred query is determined based on a first set of context signals associated with a triggering event. Items matching the inferred query are selected based on a second set of one or more context signals associated with the triggering event. One or more results items identifying the one or more matching items are caused to display. Result items may be displayed in association with an explanation that identifies context values used to select an item or one or more attributes of the item that are relevant to a user or the inferred query.
1. A computer-implemented method, comprising:
maintaining an affinity graph that represents a relationship between a user of a first computer device and each of a list of items of a first type,
the relationship between the user and a specific item of the list of items being related to how often the user accesses the specific item or how much the user pays to access the specific item;
detecting, by a processor, an occurrence of a triggering event for providing a recommendation for items of the first type to the first computer device,
the triggering event being associated with a first set of context signals;
in response to the triggering event, causing, by the processor, transmission of a request to the first computer device for a second set of context signals currently gathered by the first computer device,
the second set of context signals including a location or an acceleration of the first computer device;
receiving, from the first computer device, the second set of context signals;
inferring a first query related to the items of the first type based on a first subset of the combination of the first set of context signals and the second set of context signals;
adding one or more items of the first type matching the first query to the recommendation based on a second subset of the combination of the first set of context signals and the second set of context signals;
causing display of the recommendation by the first computer device.
2. The computer-implemented method of claim 1 , the second set of context signals including an attribute of the first type of items provided by a user of the first computer.
3. The computer-implemented method of claim 1 , the triggering event being that at least the first set of context signals have been collected by the processor.
4. The computer-implemented method of claim 1 , the triggering event being that a recommendation application has been launched by the first computer device.
5. The computer-implemented method of claim 1 , further comprising:
automatically inferring a second query related to items of a second type for the one item added to the recommendation;
adding one or more items of the second type matching the second inferred query to the recommendation based on a third subset of the combination of the first set of context signals and the second set of context signals.
6. The computer-implemented method of claim 5 ,
the one or more items of the first type being restaurants,
the one or more items of the second type being dishes served by the one restaurant added to the recommendation.
7. The computer-implemented method of claim 1 , further comprising:
causing display of the first query by the first computer device for a certain period of time;
causing display of the recommendation by the first computer device at the end of the certain period of time.
8. The computer-implemented method of claim 1 , the first set of context signals including a selection of an item of the first type in a prior recommendation transmitted to the first computer device.
9. The computer-implemented method of claim 1 , the first set of context signals including a specific context signal received from a second computer device within a certain time from when the request is transmitted to the first computer device.
10. A non-transitory computer-readable medium storing one or more instructions which, when executed by one or more processors, cause the one or more processors to perform a method, the method comprising:
maintaining an affinity graph that represents a relationship between a user of a first computer device and each of a list of items of a first type,
the relationship between the user and a specific item of the list of items being related to how often the user accesses the specific item or how much the user pays to access the specific item;
detecting an occurrence of a triggering event for providing a recommendation for items of the first type to the first computer device, the triggering event being associated with a first set of context signals;
in response to the triggering event, causing transmission of a request to the first computer device for a second set of context signals currently gathered by the first computer device,
the second set of context signals including a location or an acceleration of the first computer device;
receiving, from the first computer device, the second set of context signals;
inferring a first query related to the items of the first type based on a first subset of the combination of the first set of context signals and the second set of context signals;
adding one or more items of the first type matching the first query to the recommendation based on a second subset of the combination of the first set of context signals and the second set of context signals;
causing display of the recommendation by the first computer device.
11. The non-transitory computer-readable medium of claim 10 , the second set of context signals including an attribute of the first type of items provided by a user of the first computer.
12. The non-transitory computer-readable medium of claim 10 , the triggering event being that at least the first set of context signals have been collected by the processor.
13. The non-transitory computer-readable medium of claim 10 , the triggering event being that a recommendation application has been launched by the first computer device.
14. The non-transitory computer-readable medium of claim 10 , the method further comprising:
automatically inferring a second query related to items of a second type for the one item added to the recommendation;
adding one or more items of the second type matching the second inferred query to the recommendation based on a third subset of the combination of the first set of context signals and the second set of context signals.
15. The non-transitory computer-readable medium of claim 14 ,
the one or more items of the first type being restaurants,
the one or more items of the second type being dishes served by the one restaurant added to the recommendation.
16. The non-transitory computer-readable medium of claim 10 , the method further comprising:
causing display of the first query by the first computer device for a certain period of time;
causing display of the recommendation by the first computer device at the end of the certain period of time.
17. The non-transitory computer-readable medium of claim 10 , the first set of context signals including a selection of an item of the first type in a prior recommendation transmitted to the first computer device.
18. The non-transitory computer-readable medium of claim 10 , the first set of context signals including a specific context signal received from a second computer device within a certain time from when the request is transmitted to the first computer device.
19. The computer-implemented method of claim 1 , at least one of the first set of context signals being based on the affinity graph.
20. The non-transitory computer-readable medium of claim 10 , at least one of the first set of context signals being based on the affinity graph.