MIT Signals & Systems

The Signals & Systems project demonstrated the power of mapping applied to a real MIT course. We parsed the course's LaTeX lecture notes into structured HTML, mapped every section and concept to learning outcomes, and built a web app that lets students search for resources by intentrather than by keyword — for example, searching “I need to understand convolution” surfaces the specific sections and outcomes that address it.

How it works

All course content is modeled as a knowledge graph: lecture sections are entities, learning outcomes are entities, and directed edges encode prerequisite and coverage relationships between them. When a student searches, the query is matched against outcome statements in the graph rather than raw text — returning semantically relevant results even when the exact words don't match.

The graph is built using the Xocesmapping library and served via API. The frontend was a React + Redux single-page app with MathJax for equation rendering. Students could also add results to a “reading list” pane to study multiple sections side by side.

Explore the knowledge map

The interactive map below shows the full Signals & Systems knowledge graph — topics, outcomes, and their prerequisite relationships across the course.

Technology

The project was built with React, Redux, and a custom LaTeX-to-HTML converter that preserved mathematical structure and generated internal cross-links between related sections. All content is fed via API — no static HTML — enabling the search and graph traversal features.

Read more about the underlying methodology in our paper on network models for mapping educational data.