Most companies have internal databases that consist of both structured and unstructured content. This environment is similar to the Web setting where the "semantic" Web contains structured, machinereadable content, and the rest of the Web contains unstructured content. In this talk, I will give examples where the structured content can be augmented (both in terms of structure and content) by applying machine learning techniques to the unstructured content for a variety of business problems. I will focus on describing techniques using supervised and semi-supervised learning algorithms to e±ciently create and populate ontologies for a variety of product categories.