The following contains codes as part of the analysis of the community structure of the IRC logs channel.
The steps in this are summarised in the photo given below:
Here is a snippet of the admin interface and the user interface:
We have attached codes for:
- Web Scraping
- Classifications into users and experts
- Testing on a custom WhatsApp chat.
- Testing on a custom Telegram chat.
The codes are organised as follows:
Web_Scraping_2.ipynb contains code that scrapes the web for the IRC chat logs.
IRC_NAMES.ipynb contains code that extracts top experts and their active timings for each keyword.
Dashboard.txt contains this output, which will be displayed on the chat-bot interface.
QSTP_Competitive.ipynb contains code that extracts the usernames and timings for a university group chat on WhatsApp.
Telegram_Scraping.ipynb contains code that extracts the usernames and timings for a university group chat on Telegram.
AdjacencyList.ipynb generates the adjacency list for the graph to be made. (Shown below)
Here is a snippet of the directed graph for various users in the Ubuntu IRC chat in 2013:
Following is the graph for users/experts with colour denoting the various communities, and nodes ranked by betweenness centrality:
