SEAS SearchKG-Based Course QA

Knowledge Graph Visualization

Interactive exploration of course relationships, prerequisites, instructors, and topics

Interactive Force-Directed Graph

Courses
Professors
Topics
Loading knowledge graph...

Click nodes to view details • Drag to pan • Scroll to zoom • Click background to reset

Graph Legend

This knowledge graph represents relationships between courses, prerequisites, instructors, and topics extracted from GWU course data.

Course Nodes
Individual courses with prerequisites and topics
Professor Nodes
Instructors teaching courses
Topic Nodes
Subject areas covered by courses
Edge Types
Prerequisites
Taught By
Covers Topic

Click on any node to see detailed information and connected relationships.

Sample Multi-Hop Queries

Examples of complex prerequisite chain questions from the KG-based training data

QUERY 1

If I finished CSCI 6531, which courses remain before CSCI 8531?

CSCI 6531 → CSCI 8531

After completing CSCI 6531, you are ready to take CSCI 8531.

QUERY 2

What courses do I need after CSCI 2461 to enroll in CSCI 3410?

CSCI 2461 → CSCI 3410

To prepare for CSCI 3410, you should also take: CSCI 1112, CSCI 1111, CSCI 2113, MATH 1221.

QUERY 3

Which courses should I take to prepare for CSCI 4345 if I've completed CSCI 4342?

CSCI 4342 → CSCI 4345

To prepare for CSCI 4345, you should also take: CSCI 2113, MATH 1221.

These queries demonstrate the multi-hop reasoning capability enabled by the knowledge graph structure.

Graph Construction Methodology

Prerequisites Extraction

Regex patterns to extract prerequisite relationships from course descriptions:

r"Prerequisites?: ([A-Z]+ \d+)"

Topic Extraction

Keyword matching + spaCy NLP to identify topics from descriptions:

machine learning, algorithms, databases, security, networks, AI, software engineering, etc.

Instructor Mapping

Direct mapping from Spring 2026 schedule data:

Course sections → Instructors creates taught_by relationships