GrEmLIn: A Repository of Green Baseline Embeddings for 87 Low-Resource Languages Injected with Multilingual Graph Knowledge

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Gurgurov, Daniil, Kumar, Rishu, Ostermann, Simon
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912204913115136
author Gurgurov, Daniil
Kumar, Rishu
Ostermann, Simon
author_facet Gurgurov, Daniil
Kumar, Rishu
Ostermann, Simon
contents Contextualized embeddings based on large language models (LLMs) are available for various languages, but their coverage is often limited for lower resourced languages. Using LLMs for such languages is often difficult due to a high computational cost; not only during training, but also during inference. Static word embeddings are much more resource-efficient ("green"), and thus still provide value, particularly for very low-resource languages. There is, however, a notable lack of comprehensive repositories with such embeddings for diverse languages. To address this gap, we present GrEmLIn, a centralized repository of green, static baseline embeddings for 87 mid- and low-resource languages. We compute GrEmLIn embeddings with a novel method that enhances GloVe embeddings by integrating multilingual graph knowledge, which makes our static embeddings competitive with LLM representations, while being parameter-free at inference time. Our experiments demonstrate that GrEmLIn embeddings outperform state-of-the-art contextualized embeddings from E5 on the task of lexical similarity. They remain competitive in extrinsic evaluation tasks like sentiment analysis and natural language inference, with average performance gaps of just 5-10\% or less compared to state-of-the-art models, given a sufficient vocabulary overlap with the target task, and underperform only on topic classification. Our code and embeddings are publicly available at https://huggingface.co/DFKI.
format Preprint
id arxiv_https___arxiv_org_abs_2409_18193
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle GrEmLIn: A Repository of Green Baseline Embeddings for 87 Low-Resource Languages Injected with Multilingual Graph Knowledge
Gurgurov, Daniil
Kumar, Rishu
Ostermann, Simon
Computation and Language
Contextualized embeddings based on large language models (LLMs) are available for various languages, but their coverage is often limited for lower resourced languages. Using LLMs for such languages is often difficult due to a high computational cost; not only during training, but also during inference. Static word embeddings are much more resource-efficient ("green"), and thus still provide value, particularly for very low-resource languages. There is, however, a notable lack of comprehensive repositories with such embeddings for diverse languages. To address this gap, we present GrEmLIn, a centralized repository of green, static baseline embeddings for 87 mid- and low-resource languages. We compute GrEmLIn embeddings with a novel method that enhances GloVe embeddings by integrating multilingual graph knowledge, which makes our static embeddings competitive with LLM representations, while being parameter-free at inference time. Our experiments demonstrate that GrEmLIn embeddings outperform state-of-the-art contextualized embeddings from E5 on the task of lexical similarity. They remain competitive in extrinsic evaluation tasks like sentiment analysis and natural language inference, with average performance gaps of just 5-10\% or less compared to state-of-the-art models, given a sufficient vocabulary overlap with the target task, and underperform only on topic classification. Our code and embeddings are publicly available at https://huggingface.co/DFKI.
title GrEmLIn: A Repository of Green Baseline Embeddings for 87 Low-Resource Languages Injected with Multilingual Graph Knowledge
topic Computation and Language
url https://arxiv.org/abs/2409.18193