Project Sentry: Project Manager With Ai Based Duplicate Topic Detection

Abstract

With The Increasing Volume Of Final-year Project Sub- Missions In Educational Institutions, Ensuring The Uniqueness And Originality Of Student Work Has Become A Pressing Issue. Duplicate Project Topics Compromise Innovation And Create Challenges For Faculty During The Evaluation Process. Traditional Methods Of Identifying Duplicates Are Manual, Time-consuming, And Prone To Human Error. To Address This Issue Effectively, There Is A Need For An Intelligent System That Not Only Detects Duplicate Topics But Also Supports Streamlined Project Management. This Research Introduces An AI-driven Approach That Leverages Natural Language Processing And Machine Learning To Detect Duplicate Project Submissions. Key Textual Components Such As Project Title, Description, Domain, And Technologies Are Combined, Preprocessed, And Transformed Into Numerical Features Using TF- IDF And Dimensionality Reduction Techniques Like SVD. A Robust Classifier, Specifically XGBoost, Is Trained To Distinguish Between Unique And Duplicate Projects. In Addition To Classification, The System Uses Cosine Similarity To Provide A Duplication Score And Prediction Confidence, Ensuring Transparency And Reliability In Decision-making. Beyond Prediction, The System Also Contributes To Efficient Project Management. By Organizing Project Data, Tracking Sub- Mission Originality, And Integrating A Duplicate Detection Mecha- Nism, The System Aids Faculty In Supervising Projects Effectively. It Empowers Academic Institutions To Uphold Integrity While Managing Large Datasets Of Student Work. The Solution Is Scalable, Interpretable, And Practical For Integration Into College-level Project Portals, Fostering A Culture Of Innovation And Account- Ability In Academic Environments.

Keywords