TEXT Open Intent Recognition (TEXTOIR)

TEXTOIR is the first high-quality Text Open Intent Recognition platform. This repo contains a convenient toolkit with extensible interfaces, integrating a series of state-of-the-art algorithms of two tasks (open intent detection and open intent discovery). We also release the pipeline framework and the visualized platform in the repo TEXTOIR-DEMO.

Introduction

TEXTOIR aims to provide a convenience toolkit for researchers to reproduce the related text open classification and clustering methods. It contains two tasks, which are defined as open intent detection and open intent discovery. Open intent detection aims to identify n-class known intents, and detect one-class open intent. Open intent discovery aims to leverage limited prior knowledge of known intents to find fine-grained known and open intent-wise clusters. Related papers and codes are collected in our previous released reading list.

Open Intent Recognition:

We strongly recommend you to use our TEXTOIR toolkit, which has standard and unified interfaces (especially data setting) to obtain fair and persuable results on benchmark intent datasets!

Benchmark Datasets

Dataset Name	Source
BANKING	Paper Dataset
OOS	Paper Dataset
StackOverflow	Paper Dataset

Integrated Models

Open Intent Detection

Model Name	Source	Published
DA-ADB	Paper Code	arXiv 2022
ARPL	Paper Code	IEEE TPAMI 2022
MDF	Paper Code	ACL 2021
(K+1)-way	Paper Code	ACL 2021
ADB	Paper Code	AAAI 2021
SEG	Paper Code	ACL 2020
DeepUnk	Paper Code	ACL 2019
DOC	Paper Code	EMNLP 2017
MSP	Paper Code	ICLR 2017
OpenMax	Paper Code	CVPR 2016

Open Intent Discovery

Setting	Model Name	Source	Published
Semi-supervised	DeepAligned	Paper Code	AAAI 2021
Semi-supervised	CDAC+	Paper Code	AAAI 2020
Semi-supervised	DTC*	Paper Code	ICCV 2019
Semi-supervised	MCL*	Paper Code	ICLR 2019
Semi-supervised	KCL*	Paper Code	ICLR 2018
Unsupervised	DCN	Paper Code	ICML 2017
Unsupervised	DEC	Paper Code	ICML 2016
Unsupervised	SAE-KM	Paper	JMLR 2010
Unsupervised	AG	Paper	PR 1978
Unsupervised	KM	Paper	1967

(* denotes the CV model replaced with the BERT backbone)

Quick Start

1. Use anaconda to create Python (version >= 3.6) environment

conda create --name textoir python=3.6
conda activate textoir

2. Install PyTorch (Cuda version 11.2)

conda install pytorch torchvision torchaudio cudatoolkit=11.0 -c pytorch -c conda-forge  

2. Clone the TEXTOIR repository, and choose the task (Take open intent detection as an example).

git clone git@github.com:thuiar/TEXTOIR.git
cd TEXTOIR
cd open_intent_detection

3. Install related environmental dependencies

pip install -r requirements.txt

4. Run examples (Take ADB as an example)

sh examples/run_ADB.sh

Extensibility

This toolkit is extensible and supports adding new methods, datasets, configurations, backbones, dataloaders, losses conveniently. More detailed information can be seen in the directory open_intent_detection and open_intent_discovery respectively.

Citations

If this work is helpful, or you want to use the codes and results in this repo, please cite the following papers:

• TEXTOIR: An Integrated and Visualized Platform for Text Open Intent Recognition
• Learning Discriminative Representations and Decision Boundaries for Open Intent Detection

@inproceedings{zhang-etal-2021-textoir,
    title = "{TEXTOIR}: An Integrated and Visualized Platform for Text Open Intent Recognition",
    author = "Zhang, Hanlei  and Li, Xiaoteng  and Xu, Hua  and Zhang, Panpan and Zhao, Kang  and Gao, Kai",
    booktitle = "Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing: System Demonstrations",
    pages = "167--174",
    year = "2021",
    url = "https://aclanthology.org/2021.acl-demo.20",
    doi = "10.18653/v1/2021.acl-demo.20",
}

@article{zhang2022towards,
  title={Learning Discriminative Representations and Decision Boundaries for Open Intent Detection},
  author={Zhang, Hanlei and Xu, Hua and Zhao, Shaojie and Zhou, Qianrui},
  journal={arXiv preprint arXiv:2203.05823},
  year={2022}
}

Contributors

Hanlei Zhang, Shaojie Zhao, Xin Wang, Ting-En Lin, Qianrui Zhou, Huisheng Mao.

Bugs or questions?

If you have any questions, please open issues and illustrate your problems as detailed as possible. If you want to integrate your method in our repo, please feel free to pull request!