Enhanced personalization in multi-user devices
View Patent ↗Disclosed herein is an enhanced device personalization that personalizes a user's experience with a device, e.g., a multi-user device. Rather than personalizing based on the specific user(s) that are using, an active entity, which represents a type of consumption, may be generated and used to make recommendations for personalizing an experience using the device. In a case of a multi-user device, each user's experience is personalized by determining an active entity for the user's experience and without knowledge of which user is using the device.
1. A method comprising:
obtaining, by at least one computing device, data about a plurality of content items consumed using a plurality of devices, at least one of the devices is a multi-user device;
generating, by the at least one computing device and using the obtained data, a plurality of entities, each entity, of the plurality of entities, representing an associated type of consumption different from a type of consumption associated with each other entity of the plurality, the type of consumption associated with each entity indicating a type of content item for consumption, each entity being identified independent of those users of a plurality of users that use the plurality of devices to consume the plurality of content items;
identifying, by the at least one computing device and independent of the user using the multi-user device, an entity of the plurality using an entity over device distribution corresponding to the multi-user device, the entity over device distribution comprising a probability of the multi-user device being used to consume one or more content items of the plurality of content items of a type represented by the entity's associated type of consumption; and
making, by the at least one computing device and using the identified entity of the plurality, a recommendation comprising one or more content items of the plurality for consumption using the multi-user device, the recommendation being made independent of information identifying which one or more users of the plurality is using the multi-user device.
2. The method of claim 1 , the generating further comprising:
generating an item probability distribution over entity comprising a probability distribution for each entity of the plurality, each entity's probability distribution comprising a probability for each content item of the plurality, each content item's probability representing a probability of consumption of the content item in a context of the entity's associated type of consumption.
3. The method of claim 1 , the generating further comprising:
generating an entity distribution over device comprising a probability distribution for each device of the plurality, each device's probability distribution comprising a probability for each entity of the plurality, each entity's probability representing a probability of the device being used to consume one or more content items of a type represented by the entity's associated type of consumption.
4. The method of claim 1 , entity identification further comprising:
identifying the entity of the plurality of entities using information about a time interval to consume the one or more content items of the recommendation and information identifying the device of the plurality to be used to consume the one or more content items of the recommendation, the identifying further using a probability distribution determined for the time interval, the probability distribution comprising a probability for each entity of the plurality, each entity's probability indicating a likelihood of the device being used for the entity's associated type of consumption for the time interval.
5. The method of claim 1 , further comprising:
generating an entity distribution over device comprising a probability distribution for each device of the plurality, each device's probability distribution comprising a probability for each entity of the plurality, each entity's probability representing a probability of the device being used to consume one or more content items of a type represented by the entity's associated type of consumption;
generating an item probability distribution over entity comprising a probability distribution for each entity of the plurality, each entity's probability distribution comprising a probability for each content item of the plurality, each content item's probability representing a probability of consumption of the content item in a context of the entity's associated type of consumption;
identifying the entity of the plurality of entities further comprising identifying the entity of the plurality of entities using information about the content item being consumed using the device of the plurality and information identifying the device of the plurality, the identifying using the entity distribution over device and the item distribution over entity.
6. The method of claim 1 , making a recommendation further comprising:
ranking the plurality of content items according to each content item's probability of belonging to the entity used to make the recommendation; and
selecting a number of content items from the plurality using the ranking.
7. The method of claim 6 , selecting further comprising:
selecting a number of content items from the plurality using the ranking and at least one filter.
8. A system comprising:
at least one computing device comprising one or more processors to execute and memory to store instructions to:
obtain data about a plurality of content items consumed using a plurality of devices, at least one of the devices is a multi-user device;
generate a plurality of entities using the obtained data, each entity, of the plurality of entities, representing an associated type of consumption different from a type of consumption associated with each other entity of the plurality, the type of consumption associated with each entity indicating a type of content item for consumption, each entity being identified independent of those users of a plurality of users that use the plurality of devices to consume the plurality of content items;
identify, independent of the user using the multi-user device, an entity of the plurality using an entity over device distribution corresponding to the multi-user device, the entity over device distribution comprising a probability of the multi-user device being used to consume one or more content items of the plurality of content items of a type represented by the entity's associated type of consumption; and
make, using the identified entity of the plurality, a recommendation comprising one or more content items of the plurality for consumption using the multi-user device, the recommendation being made independent of information identifying which one or more users of the plurality is using the multi-user device.
9. The system of claim 8 , the instructions to generate further comprising instructions to:
generate an item probability distribution over entity comprising a probability distribution for each entity of the plurality, each entity's probability distribution comprising a probability for each content item of the plurality, each content item's probability representing a probability of consumption of the content item in a context of the entity's associated type of consumption.
10. The system of claim 8 , the instructions to generate further comprising instructions to:
generate an entity distribution over device comprising a probability distribution for each device of the plurality, each device's probability distribution comprising a probability for each entity of the plurality, each entity's probability representing a probability of the device being used to consume one or more content items of a type represented by the entity's associated type of consumption.
11. The system of claim 8 , the instructions to make a recommendation further comprising instructions to:
identify the entity of the plurality of entities using information about a time interval to consume the one or more content items of the recommendation and information identifying the device of the plurality to be used to consume the one or more content items of the recommendation, the identifying further using a probability distribution determined for the time interval, the probability distribution comprising a probability for each entity of the plurality, each entity's probability indicating a likelihood of the device being used for the entity's associated type of consumption for the time interval.
12. The system of claim 8 , the instructions to make a recommendation further comprising instructions to:
generate an entity distribution over device comprising a probability distribution for each device of the plurality, each device's probability distribution comprising a probability for each entity of the plurality, each entity's probability representing a probability of the device being used to consume one or more content items of a type represented by the entity's associated type of consumption;
generate an item probability distribution over entity comprising a probability distribution for each entity of the plurality, each entity's probability distribution comprising a probability for each content item of the plurality, each content item's probability representing a probability of consumption of the content item in a context of the entity's associated type of consumption;
identify the entity of the plurality of entities using information about the content item being consumed using the device of the plurality and information identifying the device of the plurality, the identifying using the entity distribution over device and the item distribution over entity.
13. The system of claim 8 , the instructions to make a recommendation further comprising instructions to:
rank the plurality of content items according to each content item's probability of belonging to the entity used to make the recommendation; and
select a number of content items from the plurality using the ranking.
14. The system of claim 13 , the instructions to select further comprising instructions to:
select a number of content items from the plurality using the ranking and at least one filter.
15. A computer readable non-transitory storage medium for tangibly storing thereon computer readable instructions that when executed cause at least one processor to:
obtain data about a plurality of content items consumed using a plurality of devices, at least one of the devices is a multi-user device;
generate a plurality of entities using the obtained data, each entity, of the plurality, representing an associated type of consumption different from a type of consumption associated with each other entity of the plurality, the type of consumption associated with each entity indicating a type of content item for consumption, each entity being identified independent of those users of a plurality of users that use the plurality of devices to consume the plurality of content items;
identify, independent of the user using the multi-user device, an entity of the plurality using an entity over device distribution corresponding to the multi-user device, the entity over device distribution comprising a probability of the multi-user device being used to consume one or more content items of the plurality of content items of a type represented by the entity's associated type of consumption; and
make, using the identified entity of the plurality, a recommendation comprising one or more content items of the plurality for consumption using the multi-user device, the recommendation being made independent of information identifying which one or more users of the plurality is using the multi-user device.
16. The medium of claim 15 , the instructions to generate further comprising instructions to:
generate an item probability distribution over entity comprising a probability distribution for each entity of the plurality, each entity's probability distribution comprising a probability for each content item of the plurality, each content item's probability representing a probability of consumption of the content item in a context of the entity's associated type of consumption.
17. The medium of claim 15 , the instructions to generate further comprising instructions to:
generate an entity distribution over device comprising a probability distribution for each device of the plurality, each device's probability distribution comprising a probability for each entity of the plurality, each entity's probability representing a probability of the device being used to consume one or more content items of a type represented by the entity's associated type of consumption.
18. The medium of claim 15 , the instructions to make a recommendation further comprising instructions to:
identify the entity of the plurality of entities using information about a time interval to consume the one or more content items of the recommendation and information identifying the device of the plurality to be used to consume the one or more content items of the recommendation, the identifying further using a probability distribution determined for the time interval, the probability distribution comprising a probability for each entity of the plurality, each entity's probability indicating a likelihood of the device being used for the entity's associated type of consumption for the time interval.
19. The medium of claim 15 , the instructions to make a recommendation further comprising instructions to:
generate an entity distribution over device comprising a probability distribution for each device of the plurality, each device's probability distribution comprising a probability for each entity of the plurality, each entity's probability representing a probability of the device being used to consume one or more content items of a type represented by the entity's associated type of consumption;
generate an item probability distribution over entity comprising a probability distribution for each entity of the plurality, each entity's probability distribution comprising a probability for each content item of the plurality, each content item's probability representing a probability of consumption of the content item in a context of the entity's associated type of consumption;
identify the entity of the plurality of entities using information about the content item being consumed using the device of the plurality and information identifying the device of the plurality, the identifying using the entity distribution over device and the item distribution over entity.
20. The medium of claim 15 , the instructions to make a recommendation further comprising instructions to:
rank the plurality of content items according to each content item's probability of belonging to the entity used to make the recommendation; and
select a number of content items from the plurality using the ranking.
21. The medium of claim 20 , the instructions to select further comprising instructions to:
select a number of content items from the plurality using the ranking and at least one filter.