Implementation of the research paper Deep Unknown Intent Detection with Margin Loss (ACL2019)

Identifying the unknown (novel) user intents that have never appeared in the training set is a challenging task in the dialogue system. In this paper, we present a two-stage method for detecting unknown intents. We use bidirectional long short-term memory (BiLSTM) network with the margin loss as the feature extractor.

With margin loss, we can learn discriminative deep features by forcing the network to maximize inter-class variance and to minimize intra-class variance. Then, we feed the feature vectors to the density-based novelty detection algorithm, local outlier factor (LOF), to detect unknown intents.

The architecture of the proposed method:

1. Install all required library

pip install -r requirements.txt

2. Get the GloVe embedding and modify the embedding_path in experiment.py/ipynb

wget http://nlp.stanford.edu/data/glove.6B.zip
unzip glove.6B.zip

3. Run the experiments by:

python experiment.py <dataset> <proportion>
python experiment.py SNIPS 50
python experiment.py ATIS 25

If you mentioned the method in your research, please cite this article:

@inproceedings{lin-xu-2019-deep,
    title = "Deep Unknown Intent Detection with Margin Loss",
    author = "Lin, Ting-En  and
      Xu, Hua",
    booktitle = "Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics",
    month = jul,
    year = "2019",
    address = "Florence, Italy",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/P19-1548",
    pages = "5491--5496",
}