Skip to content

README.md

Latest commit

 

History

History
137 lines (87 loc) · 3.13 KB

README.md

File metadata and controls

137 lines (87 loc) · 3.13 KB
Raw

Purpose

Our bot tracks the mood of your Slack channel in real time, and identifies for you where coworkers may not be particularly friendly. We are using a sentiment analysis algorithm from https://algorithmia.com, and each message's sentiment is rated according to that algorithm.

Setup

First, clone the respository and enter it.

$ git clone <url>
$ cd SlackSentiment

Set up the virtual environment (with python 2.7!) and config file.

virtualenv -p /usr/bin/python2.7 .
source bin/activate
pip install -r requirements.txt
cp python-rtmbot-master/rtmbot.conf.example python-rtmbot-master/rtmbot.conf

Create a Slack Bot and get the API token here.

https://YOURWORKSPACEHERE.slack.com/apps/build/custom-integration

Create an Algorithmia account and get an API token here.

https://algorithmia.com/

Update python-rtmbot-master/rtmbot.conf with your API keys from Slack and Algorithmia.

DEBUG: True

SLACK_TOKEN: "YOUR_SLACK_TOKEN"
ALGORITHMIA_KEY: "YOUR_ALGORITHMIA_KEY"

Enter the python-rtmbot-master directory and run the rtmbot.

$ cd python-rtmbot-master
$ python rtmbot.py

Check your Slack channel, and your bot will appear online.

Usage

The results from the analysis are in the form of:

  • Negative
  • Neutral
  • Postitive

For more detail check out the algorithm's page https://algorithmia.com/algorithms/nlp/SocialSentimentAnalysis.

Bot Commands

Currently there are three simple commands in the slack chatbox:

NOTE: commands are case sensitive.

NOTE: Slack bots can be invited to other channels and function as long as the files are running.

First,

current mood?

This command will display the current averages:

marvin BOT [7:45 PM]
Positive: 20.0%
Neutral: 60.0%
Negative: 20.0%

Second,

bot function?

This command will display what this bot is used for such as:

this bot analyzes sentiments in the group

Third,

show graph?

This command will display current averages as mentioned above with visuals such as a pie graph. e.g

Historical Analysis

Provided in the historical directory is a script that will review the last 24 hours of your Slack channel.

To run the historical analysis first you need your channel id. Running the following to get a list from your checkout directory.

$ python historical/channel_review.py -l

Channel Name    Slack ID
--------------  ----------
home            C0MUSFS49
random          C0ASDF34R
general         C056STSWB

Copy the id for the required channel and do the following:

$ python historical/channel_review.py -c C056STSWB
  %positive    %neutral    %negative
-----------  ----------  -----------
         30          48           22

Automated Warnings

When running the slack bot if a person types a sensitive or foul language in the chat box such as: "bloody hell" an automated message will pop up in the chat box:

please refrain from using sensitive words!

For list of all foul/senesitive keywords please refer to the bad_list.txt file in the python-rtmbot-master directory. These words comprehenisve list of common sensitive words compiled by Google.