Knowledge Infused Learning

Knowledge-infused learning enhances machine learning models by integrating structured knowledge, such as knowledge graphs, to improve performance and interpretability. Current research focuses on applying this approach to diverse domains, including healthcare (e.g., drug discovery, mental health assessment), marketing, and dialogue systems, often leveraging graph neural networks to incorporate knowledge effectively. This methodology addresses limitations of traditional deep learning by improving model accuracy, generating more explainable outputs (especially crucial in high-stakes applications like healthcare), and facilitating knowledge transfer across different data domains.