Vertical classification of web pages for structured data extraction

We propose a general hierarchical vertical classification framework, which can automatically discover the inherent hierarchical structure of relationships among verticals based on flat datasets, and then build a hierarchical classifier. We conducted a set of comparison experiments to verify the performance of it, such as with flat vs hierarchical structure of relationships, as well as among different feature selection and classification methods. Experimental results show that the hierarchical classifiers built on the basis of the proposed framework make big improvements over the flat classifiers when classifying unseen web pages. Among them, the Support Vector Machine using Odds Ratio to select discriminative features performs best.