On Horizontal and Vertical Separation in Hierarchical Text Classification

Our paper "On Horizontal and Vertical Separation in Hierarchical Text Classification", with Hosein Azarbonyad, Jaap Kamps, and Maarten Marx, has been accepted as a long paper at The ACM International Conference on the Theory of Information Retrieval (ICTIR'16). \o/ Hierarchy is an effective and common way of representing information and many real-world textual data can […]

From Probability Ranking Principle to Strong Separation Principle

Separability is a highly desirable property for constructing and operating autonomous information systems, and especially classifiers. In this post, I present a step by step argument which shows that based on the classification principles, having better separability in the feature space leads to better accuracy in the classification results. Based on the Probability Ranking Principle […]