A Cross Language Transfer Learning Algorithm for French Corpus Based on Knowledge Distillation

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F25%3ACSDIHQ9L" target="_blank" >RIV/00216208:11320/25:CSDIHQ9L - isvavai.cz</a>

Výsledek na webu

<a href="https://www.scopus.com/inward/record.uri?eid=2-s2.0-85195266478&doi=10.1109%2fEDPEE61724.2024.00154&partnerID=40&md5=9045278a9896e8c04f0f44e6af744c72" target="_blank" >https://www.scopus.com/inward/record.uri?eid=2-s2.0-85195266478&doi=10.1109%2fEDPEE61724.2024.00154&partnerID=40&md5=9045278a9896e8c04f0f44e6af744c72</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/EDPEE61724.2024.00154" target="_blank" >10.1109/EDPEE61724.2024.00154</a>

Jazyk výsledku

angličtina

Název v původním jazyce

A Cross Language Transfer Learning Algorithm for French Corpus Based on Knowledge Distillation

Popis výsledku v původním jazyce

With the deepening growth of globalization, language barriers have become a major challenge in information exchange. Cross Language Transfer (CLT) learning aims to address this issue by enabling machines to understand and generate texts in different languages. Transfer learning is currently a hot research field in machine learning (ML), which utilizes source domain knowledge related to the target domain to assist in learning the target domain. CLT aims to learn corresponding tasks in the target language using annotated samples from the source language, and is an important way to solve the problem of insufficient labeled data in the target language. and is an important way to solve the problem of insufficient labeled data in the target language. As one of the internationally recognized languages, studying CLT learning algorithms based on French is of great significance. Knowledge distillation (KD) is a method of transferring knowledge from one model to another, which has achieved great success in challenging transfer learning tasks. This article designs a KD based French corpus CLT learning algorithm, which transfers knowledge from the teacher model to the student model and introduces corpora from other languages for CLT learning. The experimental results show that the algorithm proposed in this paper has significant advantages in CLT learning and provides an effective solution for solving language barriers. © 2024 IEEE.

Název v anglickém jazyce

A Cross Language Transfer Learning Algorithm for French Corpus Based on Knowledge Distillation

Popis výsledku anglicky

With the deepening growth of globalization, language barriers have become a major challenge in information exchange. Cross Language Transfer (CLT) learning aims to address this issue by enabling machines to understand and generate texts in different languages. Transfer learning is currently a hot research field in machine learning (ML), which utilizes source domain knowledge related to the target domain to assist in learning the target domain. CLT aims to learn corresponding tasks in the target language using annotated samples from the source language, and is an important way to solve the problem of insufficient labeled data in the target language. and is an important way to solve the problem of insufficient labeled data in the target language. As one of the internationally recognized languages, studying CLT learning algorithms based on French is of great significance. Knowledge distillation (KD) is a method of transferring knowledge from one model to another, which has achieved great success in challenging transfer learning tasks. This article designs a KD based French corpus CLT learning algorithm, which transfers knowledge from the teacher model to the student model and introduces corpora from other languages for CLT learning. The experimental results show that the algorithm proposed in this paper has significant advantages in CLT learning and provides an effective solution for solving language barriers. © 2024 IEEE.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

—

Rok uplatnění

2024

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proc. - Int. Conf. Electr. Drives, Power Electron. Eng., EDPEE

ISBN

979-835039563-1

ISSN

—

e-ISSN

—

Počet stran výsledku

6

Strana od-do

801-806

Název nakladatele

Institute of Electrical and Electronics Engineers Inc.

Místo vydání

—

Místo konání akce

Athens

Datum konání akce

1. 1. 2025

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—