DravidianCodeMix: sentiment analysis and offensive language identification dataset for Dravidian languages in code-mixed text

• Published in: Language resources and evaluation Vol. 56; no. 3; pp. 765 - 806
• Main Authors: Chakravarthi, Bharathi Raja, Priyadharshini, Ruba, Muralidaran, Vigneshwaran, Jose, Navya, Suryawanshi, Shardul, Sherly, Elizabeth, McCrae, John P.
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01-09-2022; Springer Nature B.V
• Subjects: Archaeology; Bilingualism; Code switching; Computational Linguistics; Computer mediated communication; Computer Science; Data mining; Datasets; Deep learning; Dravidian languages; English language; Identification; Kannada language; Language; Language and Literature; Language identification; Languages; Linguistics; Machine learning; Malayalam; Multilingualism; Natural language; Original Paper; Presidential elections; Product reviews; Reputation management; Scripts; Sentiment analysis; Social media; Social networks; Social Sciences; Tamil language; User generated content; Verbal communication; Dravidian languages; Code-mixed; Sentiment analysis; Malayalam; Corpora; Kannada; Offensive language identification; Tamil
• ISSN: 1574-020X; 1574-0218
• DOI: 10.1007/s10579-022-09583-7

This paper describes the development of a multilingual, manually annotated dataset for three under-resourced Dravidian languages generated from social media comments. The dataset was annotated for sentiment analysis and offensive language identification for a total of more than 60,000 YouTube comments. The dataset consists of around 44,000 comments in Tamil-English, around 7000 comments in Kannada-English, and around 20,000 comments in Malayalam-English. The data was manually annotated by volunteer annotators and has a high inter-annotator agreement in Krippendorff’s alpha. The dataset contains all types of code-mixing phenomena since it comprises user-generated content from a multilingual country. We also present baseline experiments to establish benchmarks on the dataset using machine learning and deep learning methods. The dataset is available on Github and Zenodo.