@inproceedings{gupta-etal-2021-unsupervised,
    title = "Unsupervised Self-Training for Sentiment Analysis of Code-Switched Data",
    author = "Gupta, Akshat  and
      Menghani, Sargam  and
      Rallabandi, Sai Krishna  and
      Black, Alan W",
    editor = "Solorio, Thamar  and
      Chen, Shuguang  and
      Black, Alan W.  and
      Diab, Mona  and
      Sitaram, Sunayana  and
      Soto, Victor  and
      Yilmaz, Emre  and
      Srinivasan, Anirudh",
    booktitle = "Proceedings of the Fifth Workshop on Computational Approaches to Linguistic Code-Switching",
    month = jun,
    year = "2021",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest_wac_2008/2021.calcs-1.13/",
    doi = "10.18653/v1/2021.calcs-1.13",
    pages = "103--112",
    abstract = "Sentiment analysis is an important task in understanding social media content like customer reviews, Twitter and Facebook feeds etc. In multilingual communities around the world, a large amount of social media text is characterized by the presence of Code-Switching. Thus, it has become important to build models that can handle code-switched data. However, annotated code-switched data is scarce and there is a need for unsupervised models and algorithms. We propose a general framework called Unsupervised Self-Training and show its applications for the specific use case of sentiment analysis of code-switched data. We use the power of pre-trained BERT models for initialization and fine-tune them in an unsupervised manner, only using pseudo labels produced by zero-shot transfer. We test our algorithm on multiple code-switched languages and provide a detailed analysis of the learning dynamics of the algorithm with the aim of answering the question - {\textquoteleft}Does our unsupervised model understand the Code-Switched languages or does it just learn its representations?'. Our unsupervised models compete well with their supervised counterparts, with their performance reaching within 1-7{\%} (weighted F1 scores) when compared to supervised models trained for a two class problem."
}