Event-oriented sentiment-annotated Twitter datasets

EventSense datasets

We make available two datasets from Twitter collected around the 53rd Thessaloniki International Film Festival (#tiff53) and the 14th Thessaloniki Documentary Festival (#tdf14). The datasets were collected using the respective hashtags during the two events and were then manually annotated with sentiment (+1, 0, -1 for positive, neutral and negative sentiment respectively). In addition, the #tiff53 dataset is also annotated with the film(s) associated with each tweet (if any). We believe that this is a real-world challenging dataset due to the fact that it comprises tweets in both English and Greek, and that the expressed sentiment is often hard to capture.

Download the data: [tiff53] [tdf14]

Acknowledgement: In case you make use of this dataset in your work, please include a reference to the following paper, where the dataset was created:

E. Schinas, S. Papadopoulos, S. Diplaris, Y. Kompatsiaris, Y. Mass, J. Herzig, and L. Boudakidis. (2013). EventSense: capturing the pulse of large-scale events by mining social media streams. In Proceedings of the 17th Panhellenic Conference on Informatics (PCI '13). ACM, New York, NY, USA, pp. 17-24