Projects
RDFRDFSSPARQLApache FusekiLangChainFAISS

Knowledge-Graph and Hybrid Retrieval for Course Enrollment

Built a Concordia course-enrollment knowledge base and natural-language QA system comparing graph, vector, and hybrid retrieval.

2024-11-01 • 1 min read
Code

Overview

Answer course-enrollment questions with a system that can combine structured academic rules and semantic search.

System design

  • Modeled courses, lectures, topics, students, and grades from CSV and JSON as RDF/RDFS
  • Served the knowledge base through Apache Fuseki
  • Generated dynamic SPARQL queries for natural-language questions
  • Added LangChain and FAISS for vector and hybrid retrieval comparisons

Results

  • Built 43,291 RDF triples across 5,258 course URIs
  • Implemented graph, FAISS vector, and hybrid retrieval paths
  • Produced a common QA interface for side-by-side evaluation