Abstract
The increasing availability of digital educational resources has transformed academic collaboration. However, academic resource sharing platforms lack intelligent validation mechanisms, often leading to duplicate or irrelevant content. This paper presents StudyNet, an AI-powered peer-matching and document validation platform designed initially for Al Akhawayn University (AUI) and later expanded to Moroccan high schools and universities. StudyNet integrates Natural Language Processing (NLP) and vector-based similarity search using LangChain, DeepSeek, and pgvector to ensure that uploaded academic materials are relevant and not duplicated. Additionally, the platform features a peer-matching system, where students specify “Mastered Courses” and “Needed Courses”, allowing intelligent matching of users based on knowledge exchange principles. We conduct benchmarking experiments comparing DeepSeek + LangChain validation with traditional keyword-based document filtering to evaluate classification accuracy. Our results demonstrate significant improvements in content relevance, duplicate prevention, and peer collaboration efficiency. The proposed system enhances academic networking while ensuring resource integrity and fairness in educational exchanges.
Details
Presentation Type
Theme
KEYWORDS
Peer Learning, AI in Education, Document Validation, NLP, Vector Databases
