L3CubeMahaSent: A Marathi Tweet-based Sentiment Analysis Dataset

TL;DR

This paper introduces L3CubeMahaSent, the first large publicly available Marathi sentiment analysis dataset derived from Twitter, along with annotation guidelines and baseline deep learning classification results.

Contribution

It provides the first major Marathi sentiment dataset, complete with annotation guidelines and baseline results, filling a critical gap in Marathi NLP resources.

Findings

Dataset contains ~16,000 tweets in three sentiment classes.

Baseline models include CNN, LSTM, ULMFiT, and BERT.

Baseline results establish a foundation for future Marathi sentiment analysis.

Abstract

Sentiment analysis is one of the most fundamental tasks in Natural Language Processing. Popular languages like English, Arabic, Russian, Mandarin, and also Indian languages such as Hindi, Bengali, Tamil have seen a significant amount of work in this area. However, the Marathi language which is the third most popular language in India still lags behind due to the absence of proper datasets. In this paper, we present the first major publicly available Marathi Sentiment Analysis Dataset - L3CubeMahaSent. It is curated using tweets extracted from various Maharashtrian personalities' Twitter accounts. Our dataset consists of ~16,000 distinct tweets classified in three broad classes viz. positive, negative, and neutral. We also present the guidelines using which we annotated the tweets. Finally, we present the statistics of our dataset and baseline classification results using CNN, LSTM, ULMFiT, and BERT-based deep learning models.