Hieroglyphic feature-based data processing

Patent metadata

Field	Value
Publication number	US-2018365529-A1
Application number	US-201715848329-A
Country	US
Kind code	A1
Filing date	Dec 20, 2017
Priority date	Jun 14, 2017
Publication date	Dec 20, 2018
Grant date	—

Abstract

A computer-implemented method and a system are proposed. According to the method, in response to receiving a character, a first representation of the character is generated by performing word embedding processing on the character. The first representation is related to context of the character. A second representation of the character is generated by performing convolutional neural network (CNN) processing on the character. The second representation is related to a hieroglyphic feature of the character. A label for the character is determined by performing recurrent neural network (RNN) processing on the first representation and the second representation. The label indicates an attribute of the character related to the context.

First claim

1 . A computer-implemented method, comprising:
in response to receiving a character, generating a first representation of the character by performing word embedding processing on the character, wherein the first representation is related to a context of the character; generating a second representation of the character by performing convolutional neural network (CNN) processing on a stroke of the character, the second representation being related to a hieroglyphic feature of the character, wherein the CNN processing comprises performing a Wubi Chinese character encoding method on the stroke, the CNN processing performed by a feed-forward CNN comprising a pooling layer, a convolutional connected layer, and a fully connected layer, wherein the CNN processing comprises performing convolution processing, sampling processing, and full-connection processing; and determining a label for the character by performing recurrent neural network (RNN) processing on the first representation and the second representation, the label indicating an attribute of the character related to the context.

Assignees

• IBM

Inventors

• Duan Dongxu
• Jiang Jian Min
• Su Zhong
• Zhang Li
• Zhao Shi Wan

Classifications

• G06V30/19173

  Classification techniques · CPC title

• G06V10/82 Primary

  using neural networks · CPC title

• G06F18/2411

  based on the proximity to a decision surface, e.g. support vector machines · CPC title

• G06N3/045

  Combinations of networks · CPC title

• G06N3/044

  Recurrent networks, e.g. Hopfield networks · CPC title