Papers by Theodorus Fransen

6 papers
Cross-lingual Sentence Embedding using Multi-Task Learning (2021.emnlp-main)

Copied to clipboard

Challenge: Existing multilingual sentence embedding models require large parallel corpora to learn efficiently, limiting their scope.
Approach: They propose a sentence embedding framework based on an unsupervised loss function . they capture semantic similarity and relatedness between sentences using a multi-task loss function.
Outcome: The proposed framework outperforms state-of-the-art methods on STS, BUCC and Tatoeba benchmarks and on a monolingual benchmark.
Unsupervised Deep Language and Dialect Identification for Short Texts (2020.coling-main)

Copied to clipboard

Challenge: Existing methods for identifying closely related short texts are unsupervised . however, performance is poor for unsupervised methods for short texts .
Approach: They propose a method which can learn sentence embeddings and cluster assignments from short texts.
Outcome: The proposed method outperforms state-of-the-art methods in supervised settings . it can learn sentence embeddings and cluster assignments from short texts .
A Multilingual Evaluation Dataset for Monolingual Word Sense Alignment (2020.lrec-1)

Copied to clipboard

Challenge: a new dataset aims to align monolingual dictionaries with a single sense level for 15 languages . this dataset covers a wide range of languages and resources .
Approach: They propose to manually align monolingual dictionaries with possible semantic relationships . they use 15 languages to create a new baseline for the task of monolingual word sense alignment .
Outcome: The proposed dataset covers 15 languages and covers the more challenging task of linking general-purpose language.
MHE: Code-Mixed Corpora for Similar Language Identification (2022.lrec-1)

Copied to clipboard

Challenge: a new corpus of code-mixed data-sets is presented for similar language identification . the data-settings are based on a more-resourced minority language, Magahi .
Approach: They propose a Magahi-Hindi-English code-mixed corpus for similar language identification . they discuss the complexity of the data-set and provide a few baselines .
Outcome: The proposed corpus provides a language id at two levels: word and sentence.
MaCmS: Magahi Code-mixed Dataset for Sentiment Analysis (2024.lrec-main)

Copied to clipboard

Challenge: Sociolinguists and psychologists have been studying these variations in the lexicons and the language from the 50's . code-mixing is a popular method for understanding people's emotions and attitudes towards various subjects, but low-resourced languages often have a mix of scripts and languages.
Approach: They introduce a new sentiment data, MaCMS, for Magahi-Hindi-English code-mixed language, where Magai is a less-resourced minority language.
Outcome: The proposed dataset is the first Magahi-Hindi-English code-mixed dataset for sentiment analysis tasks.
Weakly-supervised Deep Cognate Detection Framework for Low-Resourced Languages Using Morphological Knowledge of Closely-Related Languages (2023.findings-emnlp)

Copied to clipboard

Challenge: Existing approaches to cognate detection focus on orthographic, phonetic or contextual models, which under-perform for most under-resourced languages.
Approach: They propose a language-agnostic weakly-supervised deep cognate detection framework for under-resourced languages using morphological knowledge from closely related languages.
Outcome: The proposed framework outperforms state-of-the-art methods on cognate detection datasets across languages and can be extended to a wide range of languages from any language family.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations