Chainer-Slack-Twitter-Dialogue
Python Shell Jupyter Notebook Makefile Ruby
Latest commit 2ea8bf0 Dec 14, 2016 @SnowMasaya update

README.md

Chainer-Slack-Twitter-Dialogue

This tool is making Dialogue Model for Slack

Thsi tool has 3 functions

1: Slack Communication 2: Learning by the Chainer 3: Collect Twitter Data

Description

This tool is making the Dialogue Model

If you see the detail about it, you see the below

http://qiita.com/GushiSnow/items/79ca7deeb976f50126d7

Install

If you don't install pyenv and virtualenv you have to install bellow

Prepare Install

linux

apt-get install pyenv 
apt-get install virtualenv 

Mac

brew install pyenv 
brew install virtualenv 
brew install homebrew/science/hdf5

Prepare Inastall2

pyenv install 3.4.1
pyenv rehash
pyenv local 3.4.1
virtualenv -p ~/.pyenv/versions/3.4.1/bin/python3.4 my_env
source my_env/bin/activate

pip install -r requirement.txt
pip install chainer=="1.5.1"

Prepare the Data

PreTrain
>[Wikipedia for Japanese]https://dumps.wikimedia.org/jawiki/latest/<br>
Train
>[Dialogue Data]https://sites.google.com/site/dialoguebreakdowndetection/<br>

SQLite

touch twitter_data.db

How to prepare the for dialogue Data

About Wikipedia Data

1: You set the wikipedia title data on the word2vec folder

2: Rename the data to jawiki-latest-all-titles-in-ns0

3: You execute the bellow script. You can get the Word2Vec Model

https://github.com/SnowMasaya/Chainer-Slack-Twitter-Dialogue/blob/master/word2vec/word2vec_execute.py

*

If you try to confirm the this code, you have to reduce the wikipedia data. The below script is the get the 5000 random data

sh random_choice.sh {Wikipedia Title data name} > {Random 5000 Choosing Wikipedia Title data name}

https://github.com/SnowMasaya/Chainer-Slack-Twitter-Dialogue/blob/master/word2vec/random_choice.sh

About Dialogue Data

1: You have to make the data folder

2: You get the Broken Dialogue corpus. And you make the file bellow

dev/〇〇.json
dev/■ ■.json
dev/◇◇◇.json

3: It is possible to split the data player_1, player_2 in the bellow script

https://github.com/SnowMasaya/Chainer-Slack-Twitter-Dialogue/blob/master/data_load.py

4: You have to split the each word in the sentence. You use mecab library. And you set the bellow data on the data folder

player_1_wakati
player_2_wakati

Prepare Twitter Key

https://apps.twitter.com/

Prepare enviroment.yml

Twitter

twitter:
    consumer_key: your consumer key 
    consumer_secret: your consumer secret
    token: your api token
    token_secret: your token secret
    mecab: your mecab dictionary

Slack

slack:
    api_token: your api token 
    channel: your channel 
    user: your user token
    mecab: your mecab dictionary 

Installing a library bellow

Requirements

    Python 3.4+
    Mecab and neolog-dict
    numpy
    chainer
    ipython
    notebook
    jinja2
    pyzmq
    tornado
    cython
    gensim
    PyYAML
    requests
    requests_oauthlib
    djehuty
    flask-slackbot
    flask
    mecab-python
    future
    websocket-client

Confirm library

ipython

Usage

Learning Chainer

*You execute python 
ipython notebook

Slack Communication

*You execute python
cd slack
python app.py

Get the Twitter Data

*You execute python
cd twitter
python twitter_get_usr_timeline.py
python sqlite_twitter.py

Code Directory Structure

Dialogue ipython notebook and Encoder Decoder Model
  - slack/     ... Slack Code
  - util/       ... Encoder Decoder tools
  - twitter/        ... Twitter Code
  - word2vec/        ... Word2Vec Code

Licence

The MIT License (MIT)

Copyright (c) 2015 Masaya Ogushi

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in
all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
THE SOFTWARE.

Author

SnowMasaya

References

Chainer
Python Slack
Chainer Machine Translation
Dialogue Data
Chainer Word2Vec
Wikipedia WordNet 日本語 Wikipedia Entity ベクトル