Intelligent Semantic Search for Academic Journals Using AI and NLP Techniques

The exponential growth of academic literature has rendered traditional keyword-based search engines increasingly inadequate for scholars seeking contextually relevant research. This study presents the design and implementation of an intelligent semantic search engine tailored for academic journals, integrating state-of-the-art Artificial Intelligence (AI) and Natural Language Processing (NLP) techniques. The proposed system leverages sentence transformer models (all-mpnet-base-v2) for semantic embeddings, enabling vector-based similarity searches, alongside spaCy for tokenization and entity recognition to enhance syntactic understanding. An ontology-based matching mechanism further aligns user queries with domain-specific research topics, while fuzzy matching and regular expressions improve error tolerance and numeric filtering (e.g., CiteScore, Impact Factor). The system architecture combines these NLP layers with Elasticsearch's hybrid search capabilities to process and rank peer-reviewed journal metadata sourced from Scopus and DOAJ. A modular FastAPI-based backend ensures scalability and responsiveness, while a lightweight frontend interface facilitates interactive user input. This research contributes a novel hybrid framework that unites neural semantic models with structured query construction, addressing limitations in current scholarly search systems. The study also introduces a benchmark methodology for evaluating semantic search performance in academic contexts, with implications for enhancing research efficiency, interdisciplinary discovery, and access to high-impact literature.