Philip Jama


Exploring relationships between concepts in a lightweight ontology, or associative network graph, through an interactive web-based application.

Project Overview

Information and knowledge can be structured in a model known as a lightweight ontology, or associative network.

This project traverses multitudes of text documents (including Wikipedia and news articles) to extract associative strength between abstract concepts. Network graphs help visualize the relationships among concept-nodes. Analyzing the connectivity between groups of nodes reveals community structure, as well as the importance (or centrality) of various concepts.

One motivation behind this project is to explore the effectiveness of representing knowledge in this graphical form. It is said that the 'third wave' of AI progress will improve the ability to abstract and contextualize information. In my opinion, graphical models will continue playing an increasingly important role in the next wave of intelligence breakthroughs.

Project Details


January 2017


Research + Development


Machine Learning, NLP, Python, Network Analysis, ReactJS