Title:
多言語語義表現モデルの訓練方法、装置、デバイス及び記憶媒体
Document Type and Number:
Japanese Patent JP7242993
Kind Code:
B2
Abstract:
The present disclosure discloses a method and apparatus for training a multilingual semantic representation model, an electronic device and a storage medium, and relates to the natural language processing field based on artificial intelligence. An implementation includes: training the multilingual semantic representation model using a plurality of training language materials represented in a plurality of languages respectively, such that the multilingual semantic representation model learns the semantic representation capability of each language; generating a corresponding mixed-language language material for each of the plurality of training language materials, the mixed-language language material including language materials in at least two languages; and training the multilingual semantic representation model using each mixed-language language material and the corresponding training language material, such that the multilingual semantic representation model learns semantic alignment information of different languages. With the technical solution of the present disclosure, the multilingual semantic representation model may learn the semantic alignment information among different languages, and then, semantic interaction among different languages may be realized based on the multilingual semantic representation model, with quite high practicability.
Inventors:
Ouyan, Shuan
Wang, Shuofuan
Sun, Yu
Wang, Shuofuan
Sun, Yu
Application Number:
JP2021114631A
Publication Date:
March 22, 2023
Filing Date:
July 09, 2021
Export Citation:
Assignee:
Beijing Baidu Netcom Science Technology Co., Ltd.
International Classes:
G06F40/30; G06F40/44; G06F40/56
Other References:
QIN, Livo et.al,CoSDA-ML: Multi-Lingual Code-Switching Data Augmentation for Zero-Shot Cross-Lingual NLP [online],2020年07月13日,pp.1-8,https://arxiv.org/pdf/2006.06402v2.pdf
YANG, Jian et.al,Alternating Language Modeling for Cross-Lingual Pre-Training [online],The Thirty-Fourth AAAI Conference on Artificial Inteligence (AAAI-20),2020年04月03日,pp.9386-9393,https://doi.org/10.1609/aaai.v34i05.6480
YANG, Jian et.al,Alternating Language Modeling for Cross-Lingual Pre-Training [online],The Thirty-Fourth AAAI Conference on Artificial Inteligence (AAAI-20),2020年04月03日,pp.9386-9393,https://doi.org/10.1609/aaai.v34i05.6480
Attorney, Agent or Firm:
Patent Attorney Corporation RYUKA International Patent Office