MLICC: A multi-label and incremental centroid-based classification of web pages by genre

Abstract

This paper proposes an improved centroid-based approach to classify web pages by genre using character n-grams extracted from URL, title, headings and anchors. To deal with the complexity of web pages and the rapid evolution of web genres, our approach implements a multi-label and incremental scheme in which web pages are classified one by one and can be affected to more than one genre. According to the similarity between the new page and each genre centroid, our approach either adjust the genre centroid or considers the new page as noise page and discards it. Conducted experiments show that our approach is very fast and achieves superior results over existing multi-label classifiers. © 2012 Springer-Verlag.