Monitoring Suspicious Behaviour on social Media using Data Mining
Project was done as a final project in a Bachelor of Science Honours Degree in Computer Science at Great Zimbabwe University in 2017.
A model built is a Twitter crawler that performs a sentiment analysis to determine whether tweets about something/someone are positive or negative wirth respect to crime behaviopr. The main goal of the project was to a create a model that collects crime related comments and tweets from Twitter, perform a sentimental analysis of of each tweet collected and to show the results by creating a webpage. The model was written with the Python programming language, uses the Twitter GET Search API to collect tweets, performs the analysis using Naive Bayes Classifier and displays the results in a simple html webpage. An additional feature was to try to put collected data in a database. This is beneficial as the model can easily fetch old data, for searches that have already been done before, and use it with the newly harvested tweets.