公开(公告)号：US10762300B1

公开(公告)日：2020-09-01

申请号：US16227673

申请日：2018-12-20

Applicant: Facebook, Inc.

Inventor： Jason E Weston ,  Antoine Bordes ,  Alexandre Lebrun ,  Martin Jean Raison

IPC: G06F40/35 ,  G06F3/0484 ,  G06F3/0482 ,  G06F16/332

Abstract: Techniques to predictively respond to user requests using natural language processing are described. In one embodiment, an apparatus may comprise a client communication component operative to receive a user service request from a user client; an interaction processing component operative to submit the user service request to a memory-based natural language processing component; generate a series of user interaction exchanges with the user client based on output from the memory-based natural language processing component, wherein the series of user interaction exchanges are represented in a memory component of the memory-based natural language processing component; and receive one or more operator instructions for the performance of the user service request from the memory-based natural language processing component; and a user interface component operative to display the one or more operator instructions in an operator console. Other embodiments are described and claimed.

公开(公告)号：US20190163801A1

公开(公告)日：2019-05-30

申请号：US15826405

申请日：2017-11-29

Applicant: Facebook, Inc.

Inventor： Adam Kal Lerer ,  Timothee Lacroix ,  Adam Joshua Fisch ,  Antoine Bordes

IPC: G06F17/30

Abstract: To generate an embedding model for entities in an online system, a first set of partitions is generated. Each partition of the first set of partitions includes a subset of entities of the online system. Each partition of at least a subset of partitions of the first set of partitions is assigned to embedding workers. Each of the embedding worker determines embedding vectors for each entity in the partition assigned to the embedding worker. A second set of partitions is generated. Each partition of at least a subset of partitions of the second set of partitions are assigned to embedding workers. Each embedding worker retrieves embedding vectors for the entities in the partition assigned to embedding worker, and determines updated embedding vectors for each of the entities based on the retrieved embedding vectors and information about interaction between the entities.

公开(公告)号：US20170277667A1

公开(公告)日：2017-09-28

申请号：US15077814

申请日：2016-03-22

Applicant: Facebook, Inc.

Inventor： Jason E. Weston ,  Antoine Bordes ,  Alexandre Lebrun ,  Martin Jean Raison

IPC: G06F17/24 ,  G06F17/27 ,  G06F3/0482

CPC classification number: G06F17/279 ,  G06F3/0482 ,  G06F3/0484 ,  G06F17/30654

Abstract: Techniques to predictively respond to user requests using natural language processing are described. In one embodiment, an apparatus may comprise a client communication component operative to receive a user service request from a user client; an interaction processing component operative to submit the user service request to a memory-based natural language processing component; generate a series of user interaction exchanges with the user client based on output from the memory-based natural language processing component, wherein the series of user interaction exchanges are represented in a memory component of the memory-based natural language processing component; and receive one or more operator instructions for the performance of the user service request from the memory-based natural language processing component; and a user interface component operative to display the one or more operator instructions in an operator console. Other embodiments are described and claimed.

公开(公告)号：US20170103324A1

公开(公告)日：2017-04-13

申请号：US14881352

申请日：2015-10-13

Applicant: Facebook, Inc.

Inventor： Jason E. Weston ,  Sumit Chopra ,  Antoine Bordes

IPC: G06N5/02 ,  G06F17/27 ,  G06F17/30

CPC classification number: G06N5/02 ,  G06F16/24578 ,  G06F16/285 ,  G06F16/90332 ,  G06N3/0445 ,  G06N3/08 ,  G06N5/04

Abstract: Embodiments are disclosed for providing a machine-generated response (e.g., answer) to an input (e.g., question) based on long-term memory information. A method according to some embodiments include receiving an input; converting the input into an input feature vector in an internal feature representation space; updating a memory data structure by incorporating the input feature vector into the memory data structure; generating an output feature vector in the internal feature representation space, based on the updated memory data structure and the input feature vector; converting the output feature vector into an output object; and providing an output based on the output object as a response to the input.

公开(公告)号：US10706074B2

公开(公告)日：2020-07-07

申请号：US15826405

申请日：2017-11-29

Applicant: Facebook, Inc.

Inventor： Adam Kal Lerer ,  Timothee Lacroix ,  Adam Joshua Fisch ,  Antoine Bordes

IPC: G06F16/27 ,  G06F16/28 ,  G06Q50/00

Abstract: To generate an embedding model for entities in an online system, a first set of partitions is generated. Each partition of the first set of partitions includes a subset of entities of the online system. Each partition of at least a subset of partitions of the first set of partitions is assigned to embedding workers. Each of the embedding worker determines embedding vectors for each entity in the partition assigned to the embedding worker. A second set of partitions is generated. Each partition of at least a subset of partitions of the second set of partitions are assigned to embedding workers. Each embedding worker retrieves embedding vectors for the entities in the partition assigned to embedding worker, and determines updated embedding vectors for each of the entities based on the retrieved embedding vectors and information about interaction between the entities.

公开(公告)号：US10198433B2

公开(公告)日：2019-02-05

申请号：US15077814

申请日：2016-03-22

Applicant: Facebook, Inc.

Inventor： Jason E Weston ,  Antoine Bordes ,  Alexandre Lebrun ,  Martin Jean Raison

IPC: G06F17/27 ,  G06F3/0482 ,  G06F3/0484 ,  G06F17/30

Abstract: Techniques to predictively respond to user requests using natural language processing are described. In one embodiment, an apparatus may comprise a client communication component operative to receive a user service request from a user client; an interaction processing component operative to submit the user service request to a memory-based natural language processing component; generate a series of user interaction exchanges with the user client based on output from the memory-based natural language processing component, wherein the series of user interaction exchanges are represented in a memory component of the memory-based natural language processing component; and receive one or more operator instructions for the performance of the user service request from the memory-based natural language processing component; and a user interface component operative to display the one or more operator instructions in an operator console. Other embodiments are described and claimed.

公开(公告)号：US20180357240A1

公开(公告)日：2018-12-13

申请号：US16002463

申请日：2018-06-07

Applicant: Facebook, Inc.

Inventor： Alexander Holden Miller ,  Adam Joshua Fisch ,  Jesse Dean Dodge ,  Amir-Hossein Karimi ,  Antoine Bordes ,  Jason E. Weston

IPC: G06F17/30 ,  G06N99/00

CPC classification number: G06F17/3053 ,  G06F17/30513 ,  G06N99/005

Abstract: In one embodiment, a computing system may generate a query vector representation of an input (e.g., a question). The system may generate relevance measures associated with a set of key-value memories based on comparisons between the query vector representation and key vector representations of the keys in the memories. The system may generate an aggregated result based on the relevance measures and value vector representations of the values in the memories. Through an iterative process that iteratively updates the query vector representation used in each iteration, the system may generate a final aggregated result using a final query vector representation. A combined feature representation may be generated based on the final aggregated result and the final query vector representation. The system may select an output (e.g., an answer to the question) in response to the input based on comparisons between the combined feature representation and a set of candidate outputs.