Dynamic and hierarchical classification of Web pages

Abstract

Automatic classification of Web pages is an effective way to organise the vast amount of information and to assist in retrieving relevant information from the Internet. Although many automatic classification systems have been proposed, most of them ignore the conflict between the fixed number of categories and the growing number of Web pages being added into the systems. They also require searching through all existing categories to make any classification. This article proposes a dynamic and hierarchical classification system that is capable of adding new categories as required, organising the Web pages into a tree structure, and classifying Web pages by searching through only one path of the tree. The proposed single-path search technique reduces the search complexity from u(n) to u(log(n)). Test results show that the system improves the accuracy of classification by 6 percent in comparison to related systems. The dynamic-category expansion technique also achieves satisfying results for adding new categories into the system as required.

Web Page Advertisement Classification

“…This is done by (1) retrieve all HTML codes of the training web page, for which we use the Mathematica function: Import[url, -Source‖]. (2) We analyzed the HTML codes to determine the feature counts, for which we use the following Mathematica code:…”

“…The proposed system can achieve reasonable average classification accuracy, which provides the groundwork for future research. We incorporated this system with our other genre [11] and subject based [1]- [5]system to create a comprehensive web page classification system [2], [4], [14].…”

Human-inspired Identification of High-level Concepts using OWA and Linguistic Quantifiers

“…Because our work is focused on a concept-identification method that involved ontology we do not describe the classifier-based techniques for identification of concepts. Some example of the work related to this topic can be found in [4], [5], [6], and [7].…”

Hierarchical Classification by Expected Utility Maximization

“…In [7], a greedy probabilistic hierarchical classifier is used to determine the most suitable path in the hierarchy from the root to a leaf. In a second step, another classifier is used to determine the best class along this path.…”

“…Roughly speaking, the lower the quality of the estimates, the more conservative the classifier should become, for example by decreasing L. To illustrate, consider a situation, in which the flat classifier is extremely misleading: It predicts a probability close to 1 for a randomly selected class and distributes the remaining probability mass uniformly over the other classes. Computing the expected utility according to (7), the optimal prediction of the hierarchical classifier will be the class with probability close to 1. However, as this class is chosen at random by the flat classifier, the true performance of the hierarchical classifier will be very poor.…”