Campus guard ai

Inspiration

Cyberbullying on social media and online platforms can affect students' confidence, mental well-being, and sense of safety. We were inspired to build CampusGuard AI to help identify potentially harmful online messages and support safer, more respectful digital communities.

What it does

CampusGuard AI is an AI-powered cyberbullying detection system that analyzes text messages and predicts whether they may contain cyberbullying. It uses Natural Language Processing (NLP) and Machine Learning to classify messages as potentially harmful or non-cyberbullying. The system aims to help users recognize harmful content earlier and encourage responsible online communication.

How we built it

We designed CampusGuard AI as a web-based project with a user-friendly interface. The proposed machine learning pipeline includes text preprocessing, TF-IDF (Term Frequency–Inverse Document Frequency) feature extraction, and a Random Forest classifier. The frontend allows users to submit text, while the backend processes the input and returns the model's prediction. We used development and debugging tools such as VS Code, with Python and relevant machine learning libraries for the AI workflow.

Challenges we ran into

One major challenge was understanding how language varies across online conversations. Slang, spelling mistakes, sarcasm, and context can make harmful messages difficult to identify accurately. Other challenges included preparing suitable labelled data, connecting the model with the web interface, handling incorrect inputs, and considering user privacy. We recognized that model predictions need careful testing and human judgement.

Accomplishments that we're proud of

We developed a practical concept that combines AI, web development, and digital safety to address a real problem affecting students. We established a clear workflow for analyzing text and presenting predictions through a simple interface. We are proud that CampusGuard AI focuses not only on technology but also on awareness, responsible online behavior, and safer campus communities.

What we learned

We learned how NLP and machine learning can be applied to real-world problems, how text preprocessing and feature extraction support classification, and why model evaluation is essential. We also learned the importance of privacy, fairness, accessibility, and human oversight when developing AI systems that affect people.

What's next for CampusGuard AI

Our future plans include improving the model with more diverse and representative datasets, supporting multiple languages, testing performance using precision, recall, and F1-score, and improving the user interface. We also plan to explore optional reporting features, a dashboard for authorized reviewers, and AWS cloud deployment. Any reporting or monitoring features will be designed with privacy protections and appropriate human review. Our goal is to help build safer, more inclusive digital communities through responsible AI.

Built With

• api