Learning to Bootstrap for Entity Set Expansion

Source code for EMNLP 2019 paper: Learning to Bootstrap for Entity Set Expansion.

Required package:

This project needs an external package provided by myself--yan_tools, you can find it here.

Preprocess:

• To accelarate, we pre-process and cache the dataset file first:

python preprocess.py

Execution

• To execute our code, please run:

python main.py --dataset "file path to your cached dataset"

Citation

Please cite the following paper if you find our code is helpful.

@inproceedings{yan-etal-2019-learning,
    title = "Learning to Bootstrap for Entity Set Expansion",
    author = "Yan, Lingyong  and
      Han, Xianpei  and
      Sun, Le  and
      He, Ben",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)",
    month = nov,
    year = "2019",
    address = "Hong Kong, China",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/D19-1028",
    doi = "10.18653/v1/D19-1028",
    pages = "292--301",
    abstract = "Bootstrapping for Entity Set Expansion (ESE) aims at iteratively acquiring new instances of a specific target category. Traditional bootstrapping methods often suffer from two problems: 1) delayed feedback, i.e., the pattern evaluation relies on both its direct extraction quality and extraction quality in later iterations. 2) sparse supervision, i.e., only few seed entities are used as the supervision. To address the above two problems, we propose a novel bootstrapping method combining the Monte Carlo Tree Search (MCTS) algorithm with a deep similarity network, which can efficiently estimate delayed feedback for pattern evaluation and adaptively score entities given sparse supervision signals. Experimental results confirm the effectiveness of the proposed method.",
}