L3Cube-HingCorpus and HingBERT: A Code Mixed Hindi-English Dataset and BERT Language Models

TL;DR

This paper introduces L3Cube-HingCorpus, a large-scale Hindi-English code-mixed dataset, and develops several transformer models including HingBERT and HingGPT, demonstrating their effectiveness on various NLP tasks and providing resources for future code-mixed language processing.

Contribution

The paper presents the first large-scale real Hindi-English code-mixed dataset and pre-trained transformer models specifically designed for code-mixed NLP tasks.

Findings

HingBERT models outperform existing models on code-mixed NLP tasks.

HingGPT can generate coherent full tweets in code-mixed language.

The datasets and models are publicly available for further research.

Abstract

Code-switching occurs when more than one language is mixed in a given sentence or a conversation. This phenomenon is more prominent on social media platforms and its adoption is increasing over time. Therefore code-mixed NLP has been extensively studied in the literature. As pre-trained transformer-based architectures are gaining popularity, we observe that real code-mixing data are scarce to pre-train large language models. We present L3Cube-HingCorpus, the first large-scale real Hindi-English code mixed data in a Roman script. It consists of 52.93M sentences and 1.04B tokens, scraped from Twitter. We further present HingBERT, HingMBERT, HingRoBERTa, and HingGPT. The BERT models have been pre-trained on codemixed HingCorpus using masked language modelling objectives. We show the effectiveness of these BERT models on the subsequent downstream tasks like code-mixed sentiment analysis, POS tagging, NER, and LID from the GLUECoS benchmark. The HingGPT is a GPT2 based generative transformer model capable of generating full tweets. We also release L3Cube-HingLID Corpus, the largest code-mixed Hindi-English language identification(LID) dataset and HingBERT-LID, a production-quality LID model to facilitate capturing of more code-mixed data using the process outlined in this work. The dataset and models are available at https://github.com/l3cube-pune/code-mixed-nlp .