Refer to “Anti-Troll_System_Using_Naive_Bayes_Algorithm/Project Code and Execution Steps/code/Firstcode/app.py” for the code
The project consists of three modules logic/registration, generic search and keyword search.
The first module is login/registration, in which you can either log in or register in our system database.
After that user can access two options: generic search and individual search.
Live tweets are fetched and evaluated using the naive bayes model trained dataset.
All the tweets are then stored in the database labeled based on whether its offensive tweet or not an offensive tweet.
Predicted report of tweets displays on the website along with the pie chart
In the individual search option , user is given option to search tweets by specific keyword.
User can input particular keywords such as person name, company name or specific brands. The tweets are then fetched according to that keyword and then labeled as either an inflammatory tweet or not based on the naive bayes classifier.
Finally dashboards, pie charts, bar graphs and scatter plots graphs are generated for showing insights to the user