PAPER ACCEPTED: Enhancing Heart Disease Prediction with Knowledge Graphs

We are excited to announce the our recent paper, “Leveraging Knowledge Graphs for Enhanced Heart Disease Prediction in Machine Learning Models”, has been accepted at the Knowledge Graphs and Neurosymbolic AI Workshop, part of the AIRoV Symposium.

In this paper, we explore the integration of Knowledge Graph (KG) embeddings with traditional tabular data to improve the accuracy and F2 score of machine learning algorithms in predicting heart disease. Our study conducts a comprehensive comparative analysis of various methodologies for merging KGs with tabular data. We compare between two different embedding algorithms in enhancing the datasets for more accurate machine learning predictions.

Our findings demonstrate an improvement in the performance of machine learning models through the incorporation of semantic information from KGs emphasizing the potential of KGs to enrich machine learning in the healthcare domain, particularly for critical applications such as heart disease prediction.

This work was supported by the FWF HOnEst project (V 754-N), FFG SENSE project (894802) and FAIR-AI project (904624).