metadata

language: en
tags:
  - roberta
  - sentiment
  - twitter
widget:
  - text: This looks tasty. Where can I buy it??
  - text: Now I want this, too.
  - text: You look great today!
  - text: I just love spring and sunshine!

This RoBERTa-based model can classify expressed purchase intentions in English language text in 2 classes:

purchase intention 🤩
no purchase intention 😐

The model was fine-tuned on 2,000 manually annotated social media posts. The hold-out accuracy is 95% (vs. a balanced 50% random-chance baseline). For details on the training approach see Web Appendix F in Hartmann et al. (2021).

Reference

Please cite this paper when you use our model. Feel free to reach out to j.p.hartmann@rug.nl with any questions or feedback you may have.

@article{hartmann2021,
  title={The Power of Brand Selfies},
  author={Hartmann, Jochen and Heitmann, Mark and Schamp, Christina and Netzer, Oded},
  journal={Journal of Marketing Research}
  year={2021}
}