Passage-Level Reading Behavior Model for Mobile Search

Thanks for visiting.

This repository contains the code and data of our paper "A Passage-Level Reading Behavior Model for Mobile Search".

The code is based on the click model project by THUIR. Please refer to this repo for detailed information on all arguments.

If you find our paper and resources useful, please kindly leave a star and cite our papers. Thanks!

@inproceedings{wu2023passage,
	author = {Wu, Zhijing and Mao, Jiaxin and Xu, Kedi and Song, Dandan and Huang, Heyan},
	title = {A Passage-Level Reading Behavior Model for Mobile Search},
	year = {2023},
	publisher = {Association for Computing Machinery},
	address = {New York, NY, USA},
	doi = {10.1145/3543507.3583343},
	booktitle = {Proceedings of the ACM Web Conference 2023},
	series = {WWW '23}
}

How to use

1. Clone or download this repo.
2. Train and test the PRM with the given data or other datasets.

python test_reading_model.py ../data ../data -m MER-VPT-V3 -o ../output/lognormal --ignore_no_clicks --ignore_no_viewport --viewport_time -V 8

• -m: The model that you would like to run.
• -o: The path to output dictionary.
• -V: The viewport time model used in PRM:
  • 8: PRM with log-normal
  • 9: PRM with gamma
  • 10: PRM with Weibull

3. Run evaluate_reading_model.py to evaluate the PRM on passage ranking and document-level relevance estimation tasks.