On this page you can find an anonymized dataset of clicks, mousing and satisfaction signals as well as crowdsourcing ratings.
This dataset is released to supplement the following publication:
- Chuklin, A. and de Rijke, M. 2016. Incorporating Clicks, Attention and Satisfaction into a Search Engine Result Page Evaluation Model. CIKM (2016).
The dataset is released under CC BY 4.0 License. In particular, under this license you are required to provide attribution, which is best achieved by citing the paper above. If your use-case makes it hard to use the dataset under such license, please get in touch with us.
Privacy and Anonymity
The original dataset contained potentially sensitive data contributed by group of volunteers (mostly from ILPS) as well as crowd workers. The privacy design doc can be found here. In preparation for the public data release, we removed as much PII as possible. We understand that it may prevent you from some additional experiments not described in the paper and encourage you to get in touch with us if you want to get some additional data.