公开(公告)号：US20180150743A1

公开(公告)日：2018-05-31

申请号：US15370305

申请日：2016-12-06

Applicant: Xerox Corporation

Inventor： Fenglong Ma ,  Radha Chitta ,  Jing Zhou ,  Palghat S. Ramesh ,  Tong Sun ,  Saurabh Singh Kataria

IPC: G06N3/08 ,  G06N3/04

CPC classification number: G06N3/084 ,  G06N3/0445 ,  G06N3/063

Abstract: A long-term memory network method and system for text comprehension. A recurrent neural network can be provided, which includes an external memory module and a long-short term memory unit, wherein said recurrent neural network encodes raw text information into vector representations, forms memories, finds relevant sentences to answer questions, and generates multi-word answers to said questions utilizing the long short term memory unit.

公开(公告)号：US10963789B2

公开(公告)日：2021-03-30

申请号：US15370305

申请日：2016-12-06

Applicant: Xerox Corporation

Inventor： Fenglong Ma ,  Radha Chitta ,  Jing Zhou ,  Palghat S. Ramesh ,  Tong Sun ,  Saurabh Singh Kataria

IPC: G06N5/02 ,  G06N5/04 ,  G06F17/27 ,  G06K9/62 ,  G06N3/08 ,  G06N3/04 ,  G06N3/063

Abstract: A long-term memory network method and system for text comprehension. A recurrent neural network can be provided, which includes an external memory module and a long-short term memory unit, wherein said recurrent neural network encodes raw text information into vector representations, forms memories, finds relevant sentences to answer questions, and generates multi-word answers to said questions utilizing the long short term memory unit.