Small sample text data hybrid enhancement method
The small sample text data hybrid enhancement method disclosed by the invention is simple, complete and high in self-adaption. The method is realized through the following technical scheme: based on a text data enhancement target, dividing an original text into long text data and short text data, au...
Saved in:
Main Authors | , , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
10.12.2021
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The small sample text data hybrid enhancement method disclosed by the invention is simple, complete and high in self-adaption. The method is realized through the following technical scheme: based on a text data enhancement target, dividing an original text into long text data and short text data, automatically separating and distinguishing the long text data and the short text data, carrying out synonym replacement, random insertion, random exchange and random deletion on the long text data, automatically adapting texts with different lengths, carrying out retranslation enhancement on the short text data, carrying out statistical analysis on text data sample length distribution, subdividing data sample distribution into groups with finer granularity, and carrying out mask prediction or pre-training; classifying each text data sample into different groups, setting different covering probabilities for the text data samples of different groups according to the groups, and performing mask prediction through a noi |
---|---|
Bibliography: | Application Number: CN202111011031 |