We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computation and Language

Title: TunBERT: Pretrained Contextualized Text Representation for Tunisian Dialect

Abstract: Pretrained contextualized text representation models learn an effective representation of a natural language to make it machine understandable. After the breakthrough of the attention mechanism, a new generation of pretrained models have been proposed achieving good performances since the introduction of the Transformer. Bidirectional Encoder Representations from Transformers (BERT) has become the state-of-the-art model for language understanding. Despite their success, most of the available models have been trained on Indo-European languages however similar research for under-represented languages and dialects remains sparse.
In this paper, we investigate the feasibility of training monolingual Transformer-based language models for under represented languages, with a specific focus on the Tunisian dialect. We evaluate our language model on sentiment analysis task, dialect identification task and reading comprehension question-answering task. We show that the use of noisy web crawled data instead of structured data (Wikipedia, articles, etc.) is more convenient for such non-standardized language. Moreover, results indicate that a relatively small web crawled dataset leads to performances that are as good as those obtained using larger datasets. Finally, our best performing TunBERT model reaches or improves the state-of-the-art in all three downstream tasks. We release the TunBERT pretrained model and the datasets used for fine-tuning.
Subjects: Computation and Language (cs.CL); Machine Learning (cs.LG)
Cite as: arXiv:2111.13138 [cs.CL]
  (or arXiv:2111.13138v1 [cs.CL] for this version)

Submission history

From: Hatem Haddad [view email]
[v1] Thu, 25 Nov 2021 15:49:50 GMT (93kb,D)

Link back to: arXiv, form interface, contact.