Extracting Knowledge About the Usage of Hate Speech from Social Media

University of Essex BSc Computer Science Final Year Project.

This project uses Social Media sites such as Twitter to extract knowledge about the usage of Hate Speech online.

The system is developed in Java 8 with Spring Boot backed by a PostgreSQL database feeding data into Grafana for monitoring and visualisation.

A deployed version of the system can be found Here.

The developed classifier was able to achieve an F1 Score of 92% on the testing dataset.

Screenshot

Grafana

Final System Architecture

Getting Started

Prerequisites

Software:

• Java 8
• Maven
• Docker
• Docker Compose

Running Application

The application runs in Docker.

The provided Docker Compose file will setup the PostgreSQL database, the Java Back End and the Grafana Front End.

cd src/java/
mvn package
cd ..
cd ..
docker-compose up

Local Application

Versioning Strategy

Using Semantic Versioning:

Given a version number MAJOR.MINOR.PATCH, increment the:

MAJOR version when you make incompatible API changes,
MINOR version when you add functionality in a backwards-compatible manner, and
PATCH version when you make backwards-compatible bug fixes.

My Minimum Viable Product is tagged as 1.0.0

Version History

1.0.0 Initial Release | Twitter Integration | Grafana Visualisation | Hate Speech Trained Classifier. - 08/12/2018.

2.0.0 Reworked API | Control Interface | Retrained Hate Speech Model. - 23/01/2019.

2.1.0 Refactor to Hate Speech from Human Rights | Confidence Measure | Improved Control Interface | Improved Classifier. - 10/04/2019

3.0.0 Refactored Whole Project to deploy using Docker-Compose. - 26/06/2019

Authors

• Ben Suskins

References

• Twitter Hosebird Client
• Spring Framework
• Apache Maven
• Stanford CoreNLP
• Grafana
• Hate Speech Training Data
• Sentiment 140 Training Data