Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/hyunwoongko/kochat
Opensource Korean chatbot framework
https://github.com/hyunwoongko/kochat
chatbot deep-learning deeplearning korean korean-chatbot sentence-classification sequance-tagging web-crawler
Last synced: 21 days ago
JSON representation
Opensource Korean chatbot framework
- Host: GitHub
- URL: https://github.com/hyunwoongko/kochat
- Owner: hyunwoongko
- License: apache-2.0
- Created: 2019-06-22T12:09:59.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2023-05-22T23:56:07.000Z (over 1 year ago)
- Last Synced: 2024-10-14T21:48:34.297Z (about 1 month ago)
- Topics: chatbot, deep-learning, deeplearning, korean, korean-chatbot, sentence-classification, sequance-tagging, web-crawler
- Language: Python
- Homepage:
- Size: 310 MB
- Stars: 450
- Watchers: 20
- Forks: 183
- Open Issues: 11
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
- awesome-nlp-note - API basd Chatbot example
README
# Kochat
[![PyPI version](https://badge.fury.io/py/kochat.svg)](https://badge.fury.io/py/kochat)
![GitHub](https://img.shields.io/github/license/gusdnd852/kochat)![introduction_kochat](https://user-images.githubusercontent.com/38183241/85958000-1b8ed080-b9cd-11ea-99d6-69b472f3e2ff.jpg)
![](https://user-images.githubusercontent.com/38183241/86410173-4347a680-bcf5-11ea-9261-e272ad21ed36.gif)
- 챗봇 빌더는 성에 안차고, 자신만의 딥러닝 챗봇 애플리케이션을 만드시고 싶으신가요?
- Kochat을 이용하면 손쉽게 자신만의 딥러닝 챗봇 애플리케이션을 빌드할 수 있습니다.```python
# 1. 데이터셋 객체 생성
dataset = Dataset(ood=True)# 2. 임베딩 프로세서 생성
emb = GensimEmbedder(model=embed.FastText())# 3. 의도(Intent) 분류기 생성
clf = DistanceClassifier(
model=intent.CNN(dataset.intent_dict),
loss=CenterLoss(dataset.intent_dict)
)# 4. 개체명(Named Entity) 인식기 생성
rcn = EntityRecognizer(
model=entity.LSTM(dataset.entity_dict),
loss=CRFLoss(dataset.entity_dict)
)# 5. 딥러닝 챗봇 RESTful API 학습 & 빌드
kochat = KochatApi(
dataset=dataset,
embed_processor=(emb, True),
intent_classifier=(clf, True),
entity_recognizer=(rcn, True),
scenarios=[
weather, dust, travel, restaurant
]
)# 6. View 소스파일과 연결
@kochat.app.route('/')
def index():
return render_template("index.html")# 7. 챗봇 애플리케이션 서버 가동
if __name__ == '__main__':
kochat.app.template_folder = kochat.root_dir + 'templates'
kochat.app.static_folder = kochat.root_dir + 'static'
kochat.app.run(port=8080, host='0.0.0.0')
```## Why Kochat?
- 한국어를 지원하는 최초의 오픈소스 딥러닝 챗봇 프레임워크입니다. (빌더와는 다릅니다.)
- 다양한 Pre built-in 모델과 Loss함수를 지원합니다. NLP를 잘 몰라도 챗봇을 만들 수 있습니다.
- 자신만의 커스텀 모델, Loss함수를 적용할 수 있습니다. NLP 전문가에겐 더욱 유용합니다.
- 챗봇에 필요한 데이터 전처리, 모델, 학습 파이프라인, RESTful API까지 모든 부분을 제공합니다.
- 가격 등을 신경쓸 필요 없으며, 앞으로도 쭉 오픈소스 프로젝트로 제공할 예정입니다.
- 아래와 같은 다양한 성능 평가 메트릭과 강력한 시각화 기능을 제공합니다.![](https://user-images.githubusercontent.com/38183241/86397184-513dfd00-bcde-11ea-9540-aa56a24b6d9b.png)
![](https://user-images.githubusercontent.com/38183241/86397411-b8f44800-bcde-11ea-8b66-22423c12584c.png)
![](https://user-images.githubusercontent.com/38183241/86396855-b47b5f80-bcdd-11ea-9672-4adf0f0ed140.png)
![](https://user-images.githubusercontent.com/38183241/86323429-c62a1c00-bc77-11ea-9caf-ede65f4cbc6c.png)
## Documentation
1. [Kochat이란?](https://github.com/gusdnd852/kochat/tree/master/docs/01_kochat_이란.md)
2. [About Chatbot](https://github.com/gusdnd852/kochat/tree/master/docs/02_about_chatbot.md)
3. [Getting Started](https://github.com/gusdnd852/kochat/tree/master/docs/03_getting_started.md)
4. [Usage](https://github.com/gusdnd852/kochat/tree/master/docs/04_usage.md)
5. [Visualization Support](https://github.com/gusdnd852/kochat/tree/master/docs/05_visualization_support.md)
6. [Performance Issue](https://github.com/gusdnd852/kochat/tree/master/docs/06_performance_issue.md)
7. [Demo](https://github.com/gusdnd852/kochat/tree/master/docs/07_demo.md)
## Reference
- [챗봇 분류 그림](https://towardsdatascience.com/chatbots-are-cool-a-framework-using-python-part-1-overview-7c69af7a7439)
- [seq2seq 그림](https://mc.ai/implement-of-seq2seq-model/)
- [Fallback Detection 그림](https://docs.smartly.ai/docs/intent-detection)
- [데모 애플리케이션 템플릿](https://bootsnipp.com/snippets/ZlkBn)
- 그 외의 그림 및 소스코드 : 본인 제작## License
```
Copyright 2020 Hyunwoong Ko.Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License athttp://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
```