Abstract:
General public of every country in the world has to go through various kinds of hardships in
their day to day life. We can name them as “public issues”. Developing countries like Sri Lanka
has a huge number of public issues. It can be a simple issue regarding a public transportation
service or a major issue regarding healthcare services, all of these fall under the same
category “Public Issues”. The purpose of this project is to create a crowdsourcing platform
named “Kiyanna” where people can report, track and rank public issues online.
The target audience of the project is the general public who has the access to internet. People
will be able to get registered with the system and post their issues. Posted issues will be listed
publicly for everyone to see. Users have the ability to up-vote or down-vote the issues listed.
The system then calculates score of each issue based on the votes received and ranks issues
accordingly.
All the issues posted on the system are analyzed and automatically classified into categories
using MALLET which is a Java based machine learning toolkit for natural languages. More
useful data about issues are extracted and visualized on a separate analytics page on the
system.
The system is operating in its beta version and available to the public at www.kiyanna.info .
System has been built on Meteor platform using Telescope framework along with MALLET for
document classification and analytics. MongoDB was used as the database. The whole system
is hosted on Amazon EC2.
