Cikm Cup 2016
Repository for the 1st place solution of the CIKM Cup 2016 (Track 1) Cross Entity Device Linking
A follow-up Work was published in proceedings of SIGIR 2017.
"Cross-Device User Linking: URL, Session, Visiting Time, and Device-log Embedding". The repository is at https://github.com/minhcp/GBUL
inproceedings{Phan:2017:CUL:3077136.3080682,
author = {Phan, Minh C. and Sun, Aixin and Tay, Yi},
title = {Cross-Device User Linking: URL, Session, Visiting Time, and Device-log Embedding},
booktitle = {Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval},
series = {SIGIR '17},
year = {2017},
isbn = {978-1-4503-5022-8},
location = {Shinjuku, Tokyo, Japan},
pages = {933--936},
numpages = {4},
url = {http://doi.acm.org/10.1145/3077136.3080682},
doi = {10.1145/3077136.3080682},
acmid = {3080682},
publisher = {ACM},
address = {New York, NY, USA},
keywords = {cross-device user linking, entity resolution},
}
}Please cite our work if you find it useful.
Actual Competition Report
A report/paper was presented. PDF
If you find this repository useful, please reference:
@misc{1610.07119,
Author = {Minh C. Phan and Yi Tay and Tuan-Anh Nguyen Pham},
Title = {Cross Device Matching for Online Advertising with Neural Feature Ensembles : First Place Solution at CIKM Cup 2016},
Year = {2016},
Eprint = {arXiv:1610.07119},
}Notes on Usage
This repository is not cleaned. There are many scripts that are experimental and not used in the end. There may be scripts for pre-processing that might be missing in this process. Use this repo at your own risk. Also, due to the space limit of git, please train the models and build the features yourself. Hopefully, more instructions should come soon.