A service-specific domain language model modeling method and a modeling system

By preprocessing scarce resource texts and analyzing them using the Gibbs algorithm, combined with the translation and relevance screening of domain-specific words, and training with an n-Gram model, the modeling challenge of specific domain language models for scarce resource languages ​​was solved, achieving efficient semantic information acquisition and improved model performance.

CN116205237BActive Publication Date: 2026-04-28INST OF ACOUSTICS CHINESE ACAD OF SCI
0 Cites 0 Cited by

Patent Information

Application Number
CN202111441264.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-11-30
Publication Date
2026-04-28
Estimated Expiration
2041-11-30

AI Technical Summary

Technical Problem

Existing language models struggle to effectively address the specific domain construction of languages ​​with scarce resources. They have low accuracy in acquiring semantic information from scarce resource corpora and cannot identify specific domain information within these corpora, resulting in poor performance for applications serving specific domains.

Method used

By collecting and preprocessing scarce resource texts, combining the Gibbs algorithm to obtain sentence-topic distribution and topic-word distribution, selecting domain feature words of the resource-rich language and translating them into the scarce language, calculating the correlation between sentences and domain topic distribution, selecting sentences with high correlation as training corpus, and using the n-Gram language model for training, a scarce resource language model is constructed.

Benefits of technology

It improves the modeling efficiency and performance of language models for languages ​​with scarce resources, accurately obtains thematic semantic information of scarce resource corpora, and enhances the processing capabilities of language models in specific domains.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The application belongs to the technical field of artificial intelligence language information processing, and particularly relates to a service specific field scarce resource language model modeling method, which comprises the following steps: step 1) obtaining a preprocessed sentence set composed of multiple segmented sentences; step 2) obtaining a sentence-topic distribution and a topic-word distribution; step 3) obtaining selected field characteristic words; step 4) obtaining translated field characteristic words; step 5) calculating and obtaining a field topic distribution by using the translated field characteristic words; step 6) calculating the correlation between the sentence-topic distribution and the field topic distribution of each segmented sentence; step 7) judging whether the correlation of each segmented sentence exceeds a preset threshold value, and selecting the sentences meeting the condition into the language model training corpus of the scarce resource specific field; step 8) obtaining the final language model training corpus; and step 9) training the scarce resource language model to obtain the scarce resource language model.
Need to check novelty before this filing date? Find Prior Art