CONNA: Addressing Name Disambiguation on the Fly

CONNA is a framework to address the problem of continues name disambiguation, a classical paper assignment task, in the current academic platforms, such as, Google Scholar, DBLP, and AMiner. More details please refer to: https://arxiv.org/abs/1910.12202 .

CONNA has been further modified accroding to the online needs, denoted as CONNA++, and deployed on AMiner which acts as the core algorithm for the effective paper assignment.

We also apply CONNA++ to a name disambiguation competition (https://www.biendata.xyz/competition/chaindream_nd_task2/) . Now, we still maintain the 1st rank in the leadboard, and we will also release the code of CONNA++ at the end of the competition.

Name Disambiguation Data

For more and cleaner name disambiguation data, we highly recommend WhoIsWho, the world largest manually-labeled name disambiguation benchmark, which contains nearly 1,000,000 papers annotated by crowdsourcing, and can be easily accessed in its dataset page,

WhoIsWho: (link: https://www.aminer.cn/whoiswho)

Citation

If you think CONNA or CONNA++ is helpful to you, please cite it as follows:

@article{chen2020conna,
  title={CONNA: Addressing Name Disambiguation on The Fly},
  author={Chen, Bo and Zhang, Jing and Tang, Jie and Cai, Lingfan and Wang, Zhaoyu and Zhao, Shu and Chen, Hong and Li, Cuiping},
  journal={IEEE Transactions on Knowledge and Data Engineering},
  year={2020},
  publisher={IEEE}
}