Інтелектуальна система рекомендацій на основі NLP-аналізу ORCID-публікацій для кафедри ші

Abstract

This research focuses on engineering an intelligent system designed to catalyze scientific collaboration within academic departments through NLP-based analysis of the ORCID metadata. The developed desktop solution leverages the .NET 8 ecosystem and the ML.NET framework to implement a robust recommendation engine grounded in TF-IDF vectorization and cosine similarity measures (S > 0.7). By autonomously identifying thematic intersections among researchers and students, the system effectively mitigates information fragmentation and facilitates the formation of an integrated, data-driven research ecosystem.