Back to results
US 11,907,272 B2Granted
Real-time personalized suggestions for communications between participants
AssigneeMicrosoft Technology Licensing, LLC
InventorsNikrouz Ghotbi, Eddie Fusaro, John Alton Price, Jeff Roger DeVries
Life of this patent
Estimating the term…
Abstract
Disclosed in some examples are methods, systems, machine-readable media, and devices which provide for real-time personalized suggestions for participants in a network-based communication service. The personalized suggestions may include options for taking actions, content suggestions, and smart replies. These suggestions may be based upon the current conversation and are delivered personally to each participant.
Claims20
A non-transitory machine readable medium that stores instructions which when performed by a machine, cause the machine to perform operations comprising:
- receiving, over a network, text of a network-based communication between a first user of a first device and a second user of a second device;
- inputting the text of the communication to a plurality of intent models, each intent model trained to detect a unique type of communicative intent in communication texts, wherein the communicative intent is a semantic meaning of the text of the communication that supports identifying suggestions associated with the following suggestion types:
- an action suggestion type, a content suggestion type, and a reply suggestion type;
- receiving a set of one or more responses from one or more of the plurality of intent models, each response in the set indicating a calculated communicative intent of the communication;
- generating a set of suggestions based upon both the set of responses and a user profile of the first user;
- ranking the set of suggestions relative to one another based upon a global model and a user model, the user model specific to preferences of the first user of the first device and customized based upon past selection history of suggestions of the first user, the global model based upon past selection history of a plurality of users;
- selecting a first suggestion and a second suggestion from the set of suggestions based upon the ranking, wherein the first suggestion is a first suggestion type and the second suggestion is a second suggestion type, the first suggestion type and the second suggestion type being different suggestion types;
- and
- sending the first suggestion and the second suggestion to the first device for presentation to the first user.
6 more elements
A computer-implemented method for providing personalized suggestions, the method comprising:
- using a computer processor:
- receiving, over a network, text of a network-based communication between a first user of a first device and a second user of a second device;
- inputting the text of the communication to a plurality of intent models, each intent model specifically trained to detect a unique type of communicative intent in communication texts, wherein the communicative intent is a semantic meaning of the text of the communications that supports identifying suggestions associated with the following suggestion types:
- an action suggestion type, a content suggestion type, and a reply suggestion type;
- receiving a set of one or more responses from one or more of the plurality of intent models, each response in the set indicating a calculated communicative intent of the communication;
- generating a set of suggestions based upon both the set of responses and a user profile of the first user;
- ranking the set of suggestions relative to one another based upon a global model and a user model, the user model specific to preferences of the first user of the first device and customized based upon past selection history of suggestions of the first user, the global model based upon past selection history of a plurality of users;
- selecting a first suggestion and a second suggestion from the set of suggestions based upon the ranking, wherein the first suggestion is a first suggestion type and the second suggestion is a second suggestion type, the first suggestion type and the second suggestion type being different suggestion types;
- and
- sending the first suggestion and the second suggestion to the first device for presentation to the first user.
7 more elements
A system for providing personalized suggestions, the system comprising:
- a processor;
- a memory, communicatively coupled to the processor and including instructions, which when performed by the processor, causes the system to perform operations comprising:
- receiving, over a network, text of a network-based communication between a first user of a first device and a second user of a second device;
- inputting the text of the communication to a plurality of intent models, each intent model trained to detect a unique type of communicative intent in communication texts, wherein the communicative intent is a semantic meaning of the text of the communications that supports identifying suggestions associated with the following suggestion types:
- an action suggestion type, a content suggestion type, and a reply suggestion type;
- receiving a set of one or more responses from one or more of the plurality of intent models, each response in the set indicating a calculated communicative intent of the communication;
- generating a set of suggestions based upon both the set of responses and a user profile of the first user;
- ranking the set of suggestions relative to one another based upon a global model and a user model, the user model specific to preferences of the first user of the first device and customized based upon past selection history of suggestions of the first user, the global model based upon past selection history of a plurality of users;
- selecting a first suggestion and a second suggestion from the set of suggestions based upon the ranking, wherein the first suggestion is a first suggestion type and the second suggestion is a second suggestion type, the first suggestion type and the second suggestion type being different suggestion types;
- and
- sending the first suggestion and the second suggestion to the first device for presentation to the first user.
8 more elements
Dependencies are inferred from the claim text.