arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2011.00677cs.CL

IndoLEM and IndoBERT: A Benchmark Dataset and Pre-trained Language Model for Indonesian NLP

The University of Melbourne（墨尔本大学）；The University of Queensland（昆士兰大学）

Fajri Koto, Afshin Rahimi, Jey Han Lau, Timothy Baldwin

备注：
Accepted at COLING 2020 - The 28th International Conference on Computational Linguistics
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

英文摘要：

Although the Indonesian language is spoken by almost 200 million people and the 10th most spoken language in the world, it is under-represented in NLP research. Previous work on Indonesian has been hampered by a lack of annotated datasets, a sparsity of language resources, and a lack of resource standardization. In this work, we release the IndoLEM dataset comprising seven tasks for the Indonesian language, spanning morpho-syntax, semantics, and discourse. We additionally release IndoBERT, a new pre-trained language model for Indonesian, and evaluate it over IndoLEM, in addition to benchmarking it against existing resources. Our experiments show that IndoBERT achieves state-of-the-art performance over most of the tasks in IndoLEM.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑