Enhancing Bilingual Lexicon Induction with Dynamic Translation

D 2025

Enhancing Bilingual Lexicon Induction with Dynamic Translation

DENISOVÁ, Michaela a Pavel RYCHLÝ

Základní údaje

Originální název

Enhancing Bilingual Lexicon Induction with Dynamic Translation

Autoři

DENISOVÁ, Michaela (703 Slovensko, garant, domácí) a Pavel RYCHLÝ (203 Česká republika, domácí)

Vydání

Porto, Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART, od s. 735-744, 10 s. 2025

Nakladatel

SciTePress

Další údaje

Jazyk

angličtina

Typ výsledku

Stať ve sborníku

Obor

10201 Computer sciences, information science, bioinformatics

Utajení

není předmětem státního či obchodního tajemství

Forma vydání

elektronická verze "online"

Organizační jednotka

Fakulta informatiky

ISBN

978-989-758-737-5

ISSN

DOI

https://doi.org/10.5220/0013346000003890

Klíčová slova anglicky

cross-Lingual embedding models; parameter k; classification neural network

Štítky

core_B, firank_B

Příznaky

Mezinárodní význam, Recenzováno

Změněno: 25. 3. 2025 08:12, Mgr. Michaela Denisová

Anotace

V originále

Bilingual lexicon induction (BLI) has been a popular task for evaluating cross-lingual word embeddings (CWEs). The prevalent metric employed in the evaluation is precision at k, where k represents the number of target words retrieved for each source word. However, establishing a fixed k for the entire evaluation dataset proves challenging due to varying target word counts for each source word. This leads to limited results, compromising either precision or recall. In this paper, we present a novel classification-based approach with dynamic k for bilingual lexicon induction that aims to identify all relevant target words for each source word by exploiting the information derived from the aligned embeddings while offering a balanced trade-off between precision and recall. On top of that, it enables the evaluation of the existing CWEs using dynamic k. Compared to the standard baseline systems and evaluation procedures, it provides competitive results.

Návaznosti

MUNI/A/1638/2024, interní kód MU

Název: Umělá inteligence a správa komplexních rozsáhlých dat

Investor: Masarykova univerzita, Umělá inteligence a správa komplexních rozsáhlých dat

Citovat

DENISOVÁ, Michaela a Pavel RYCHLÝ. Enhancing Bilingual Lexicon Induction with Dynamic Translation. Online. In Ana Paula Rocha, Luc Steels, H. Jaap van den Herik. Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART. Porto: SciTePress, 2025, s. 735-744. ISBN 978-989-758-737-5. Dostupné z: https://doi.org/10.5220/0013346000003890.

@inproceedings{2480594,
   author = {Denisová, Michaela and Rychlý, Pavel},
   address = {Porto},
   booktitle = {Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART},
   doi = {https://doi.org/10.5220/0013346000003890},
   editor = {Ana Paula Rocha, Luc Steels, H. Jaap van den Herik},
   keywords = {cross-Lingual embedding models; parameter k; classification neural network},
   howpublished = {elektronická verze "online"},
   language = {eng},
   location = {Porto},
   isbn = {978-989-758-737-5},
   pages = {735-744},
   publisher = {SciTePress},
   title = {Enhancing Bilingual Lexicon Induction with Dynamic Translation},
   year = {2025}
}

TY  - CONF
ID  - 2480594
AU  - Denisová, Michaela - Rychlý, Pavel
PY  - 2025
TI  - Enhancing Bilingual Lexicon Induction with Dynamic Translation
PB  - SciTePress
CY  - Porto
SN  - 9789897587375
KW  - cross-Lingual embedding models
KW  - parameter k
KW  - classification neural network
N2  - Bilingual lexicon induction (BLI) has been a popular task for evaluating cross-lingual word embeddings (CWEs). The prevalent metric employed in the evaluation is precision at k, where k represents the number of target words retrieved for each source word. However, establishing a fixed k for the entire evaluation dataset proves challenging due to varying target word counts for each source word. This leads to limited results, compromising either precision or recall. In this paper, we present a novel classification-based approach with dynamic k for bilingual lexicon induction that aims to identify all relevant target words for each source word by exploiting the information derived from the aligned embeddings while offering a balanced trade-off between precision and recall. On top of that, it enables the evaluation of the existing CWEs using dynamic k. Compared to the standard baseline systems and evaluation procedures, it provides competitive results.
ER  -

DENISOVÁ, Michaela a Pavel RYCHLÝ. Enhancing Bilingual Lexicon Induction with Dynamic Translation. Online. In Ana Paula Rocha, Luc Steels, H. Jaap van den Herik. \textit{Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART}. Porto: SciTePress, 2025, s.~735-744. ISBN~978-989-758-737-5. Dostupné z: https://doi.org/10.5220/0013346000003890.

Přehled o publikaci