@inproceedings{patil-etal-2023-sadtech,
    title = "{SADT}ech@{D}ravidian{L}ang{T}ech: Multimodal Sentiment Analysis of {T}amil and {M}alayalam",
    author = "Patil, Abhinav  and
      Briggs, Sam  and
      Wueger, Tara  and
      O{'}Connell, Daniel D.",
    editor = "Chakravarthi, Bharathi R.  and
      Priyadharshini, Ruba  and
      M, Anand Kumar  and
      Thavareesan, Sajeetha  and
      Sherly, Elizabeth",
    booktitle = "Proceedings of the Third Workshop on Speech and Language Technologies for Dravidian Languages",
    month = sep,
    year = "2023",
    address = "Varna, Bulgaria",
    publisher = "INCOMA Ltd., Shoumen, Bulgaria",
    url = "https://preview.aclanthology.org/add-emnlp-2024-awards/2023.dravidianlangtech-1.37/",
    pages = "250--257",
    abstract = "We present several models for sentiment analysis of multimodal movie reviews in Tamil and Malayalam into 5 separate classes: highly negative, negative, neutral, positive, and highly positive, based on the shared task, {\textquotedblleft}Multimodal Abusive Language Detection and Sentiment Analysis{\textquotedblright} at RANLP-2023. We use transformer language models to build text and audio embeddings and then compare the performance of multiple classifier models trained on these embeddings: a Multinomial Naive Bayes baseline, a Logistic Regression, a Random Forest, and an SVM. To account for class imbalance, we use both naive resampling and SMOTE. We found that without resampling, the baseline models have the same performance as a naive Majority Class Classifier. However, with resampling, logistic regression and random forest both demonstrate gains over the baseline."
}