All Projects → dialogtekgeek → DSTC6-End-to-End-Conversation-Modeling

dialogtekgeek / DSTC6-End-to-End-Conversation-Modeling

Licence: MIT license
DSTC6: End-to-End Conversation Modeling Track

Programming Languages

python
139335 projects - #7 most used programming language
shell
77523 projects

Projects that are alternatives of or similar to DSTC6-End-to-End-Conversation-Modeling

Chatbot Watson Android
An Android ChatBot powered by Watson Services - Assistant, Speech-to-Text and Text-to-Speech on IBM Cloud.
Stars: ✭ 169 (+201.79%)
Mutual labels:  chatbot, dialog, conversation
Watbot
An Android ChatBot powered by IBM Watson Services (Assistant V1, Text-to-Speech, and Speech-to-Text with Speaker Recognition) on IBM Cloud.
Stars: ✭ 64 (+14.29%)
Mutual labels:  chatbot, dialog, conversation
Seq2seqchatbots
A wrapper around tensor2tensor to flexibly train, interact, and generate data for neural chatbots.
Stars: ✭ 466 (+732.14%)
Mutual labels:  chatbot, dialog, conversation
Conversation Tensorflow
TensorFlow implementation of Conversation Models
Stars: ✭ 143 (+155.36%)
Mutual labels:  chatbot, conversation
Chatterbot
ChatterBot is a machine learning, conversational dialog engine for creating chat bots
Stars: ✭ 11,813 (+20994.64%)
Mutual labels:  chatbot, conversation
Gossiping Chinese Corpus
PTT 八卦版問答中文語料
Stars: ✭ 137 (+144.64%)
Mutual labels:  chatbot, dialog
Eddi
Scalable Open Source Chatbot Platform. Build multiple Chatbots with NLP, Behavior Rules, API Connector, Templating. Developed in Java, provided with Docker, orchestrated with Kubernetes or Openshift.
Stars: ✭ 171 (+205.36%)
Mutual labels:  chatbot, dialog
Rus-SpeechRecognition-LSTM-CTC-VoxForge
Распознавание речи русского языка используя Tensorflow, обучаясь на базе Voxforge
Stars: ✭ 50 (-10.71%)
Mutual labels:  end-to-end, lstm
Neuraldialog Cvae
Tensorflow Implementation of Knowledge-Guided CVAE for dialog generation ACL 2017. It is released by Tiancheng Zhao (Tony) from Dialog Research Center, LTI, CMU
Stars: ✭ 279 (+398.21%)
Mutual labels:  end-to-end, chatbot
Chainer Rnn Ner
Named Entity Recognition with RNN, implemented by Chainer
Stars: ✭ 19 (-66.07%)
Mutual labels:  chainer, lstm
Espnet
End-to-End Speech Processing Toolkit
Stars: ✭ 4,533 (+7994.64%)
Mutual labels:  chainer, end-to-end
Eend
End-to-End Neural Diarization
Stars: ✭ 153 (+173.21%)
Mutual labels:  chainer, end-to-end
Chinese Chatbot
中文聊天机器人,基于10万组对白训练而成,采用注意力机制,对一般问题都会生成一个有意义的答复。已上传模型,可直接运行,跑不起来直播吃键盘。
Stars: ✭ 124 (+121.43%)
Mutual labels:  chatbot, lstm
Dialog corpus
用于训练中英文对话系统的语料库 Datasets for Training Chatbot System
Stars: ✭ 1,662 (+2867.86%)
Mutual labels:  chatbot, dialog
Convform
A jQuery plugin that transforms a form into an interactive chat.
Stars: ✭ 141 (+151.79%)
Mutual labels:  chatbot, conversation
Nlp Models Tensorflow
Gathers machine learning and Tensorflow deep learning models for NLP problems, 1.13 < Tensorflow < 2.0
Stars: ✭ 1,603 (+2762.5%)
Mutual labels:  chatbot, lstm
Ai Chatbot Framework
A python chatbot framework with Natural Language Understanding and Artificial Intelligence.
Stars: ✭ 1,564 (+2692.86%)
Mutual labels:  chatbot, conversation
Xpersona
XPersona: Evaluating Multilingual Personalized Chatbot
Stars: ✭ 54 (-3.57%)
Mutual labels:  chatbot, dialog
Talquei
🤖 Vue components to build webforms looking like a conversation
Stars: ✭ 90 (+60.71%)
Mutual labels:  chatbot, conversation
Botpress
🤖 Dev tools to reliably understand text and automate conversations. Built-in NLU. Connect & deploy on any messaging channel (Slack, MS Teams, website, Telegram, etc).
Stars: ✭ 9,486 (+16839.29%)
Mutual labels:  chatbot, conversation

DSTC6: End-to-End Conversation Modeling Track

Registration

Please register: https://goo.gl/forms/Fxy061gHuSOZGC1i2

News

  • Evaluation analysis package: Jan 19 2018

    The package includes all references generated by 11 humans, hypotheses of 20 systems, and evaluation results in DSTC6 end-to-end conversation modeling track. https://www.dropbox.com/s/oh1trbos0tjzn7t/dstc6_t2_evaluation.tgz

  • Download the official training data: Sep 7-18 2017

  • Test data distribution: Sep 25 2017

  • Submission: Oct 8 2017

Easy 3 Step Data Collection

Track Description

  1. Main task (mandatory): Customer service dialog using Twitter

    (*) The tools to download the twitter data and transform to the dialog format from the data are provided.

    Task A: Full or part of the training data will be used to train conversation models.

    Task B: Any open data, e.g. from web, are available as external knowledge to generate informative sentences. But they should not overlap with the training, validation and test data provided by organizers.

  2. Pilot task: Movie scenario dialog using OpenSubtitle

  • Please cite the following paper if you will publish the results using this setup:

    https://arxiv.org/pdf/1706.07440.pdf

    @article{DSTC6_End-to-End_Conversation_Modeling,
      Author = {Chiori Hori and Takaaki Hori},
      Title = {End-to-end Conversation Modeling Track in DSTC6},    
      Journal = {arXiv:1706.07440},    
      Year = {2017}
    }
    

Necessary steps

Preparation

Most tools are written in python, which were tested on python2.7.6+ and python3.4.1+, and some bash scripts are also used to execute those tools.

For data preparation, you will need additional python modules as follows:

  • six
  • tqdm
  • nltk

which can be installed by

pip install <module-name>

or

pip install <module-name> -t <some-directory>

where <some-directory> is a directory storing python modules and needs to be accessible from python, e.g. by including it in PYTHONPATH environment variable.

If you try the baseline system, you will need Chainer http://chainer.org ,a deep learning toolkit, to perform training and evaluation of neural conversation models. Please follow the instruction in ChatbotBaseline/README.md.

Twitter task

  1. prepare data set using collect_twitter_dialogs scripts.

    $ cd collect_twitter_dialogs
    $ collect.sh
    

    (a twitter account and access keys are necessary to run the script. follow the instruction in collect_twitter_dialogs/README.md)

  2. extract training, development and test sets from stored twitter dialog data

    $ cd ../tasks/twitter
    $ make_trial_data.sh
    

    Note: the extracted data are trial data at this moment.

  3. run baseline system (optional)

    $ cd ../../ChatbotBaseline/egs/twitter
    $ run.sh
    

    (see ChatbotBaseline/README.md)

OpenSubtitles task

  1. download OpenSubtitles2016 data

    $ cd tasks/opensubs
    $ wget http://opus.lingfil.uu.se/download.php?f=OpenSubtitles2016/en.tar.gz
    $ tar zxvf en.tar.gz
    
  2. extract training, development and test sets from stored subtitle data

    $ make_trial_data.sh
    

    Note: the extracted data are trial data at this moment.

  3. run baseline system (optional)

    $ cd ../../ChatbotBaseline/egs/opensubs
    $ run.sh
    

    (see ChatbotBaseline/README.md)

Directories and files

  • README.md : this file
  • tasks : data preparation for each subtask
  • collect_twitter_dialogs : scripts to collect twitter data
  • ChatbotBaseline : a neural conversation model baseline system

Contact Information

You can get the latest updates and participate in discussions on DSTC mailing list

To join the mailing list, send an email to: ([email protected]) putting "subscribe DSTC" in the body of the message (without the quotes). To post a message, send your message to: ([email protected]).

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].