Combining ILP with Semi-supervised Learning for Web Page Categorization
نویسندگان
چکیده
This paper presents a semi-supervised learning algorithm called Iterative-Cross Training (ICT) to solve the Web pages classification problems. We apply Inductive logic programming (ILP) as a strong learner in ICT. The objective of this research is to evaluate the potential of the strong learner in order to boost the performance of the weak learner of ICT. We compare the result with the supervised Naive Bayes, which is the well-known algorithm for the text classification problem. The performance of our learning algorithm is also compare with other semi-supervised learning algorithms which are Co-Training and EM. The experimental results show that ICT algorithm outperforms those algorithms and the performance of the weak learner can be enhanced by ILP system. Keywords—Inductive Logic Programming, Semi-supervised Learning, Web Page Categorization.
منابع مشابه
Comparativa de Aproximaciones a SVM Semisupervisado Multiclase para Clasificación de Páginas Web
In this paper we present a study for semi-supervised multiclass web page classification using SVM. We propose not only combining binary semi-supervised classifiers, but also multiclass supervised ones. Our experiments show great performance for the latter method, where ignoring unlabeled documents could be better for some cases, using only labeled documents for the learning task, directly based...
متن کاملWeb Page Classification Based on Uncorrelated Semi-Supervised Intra-View and Inter-View Manifold Discriminant Feature Extraction
Web page classification has attracted increasing research interest. It is intrinsically a multi-view and semi-supervised application, since web pages usually contain two or more types of data, such as text, hyperlinks and images, and unlabeled pages are generally much more than labeled ones. Web page data is commonly high-dimensional. Thus, how to extract useful features from this kind of data ...
متن کاملThe information regularization framework for semi-supervised learning
In recent years, the study of classification shifted to algorithms for training the classifier from data that may be missing the class label. While traditional supervised classifiers already have the ability to cope with some incomplete data, the new type of classifiers do not view unlabeled data as an anomaly, and can learn from data sets in which the large majority of training points are unla...
متن کاملBoosting for multiclass semi-supervised learning
Supervised learning methods are effective when there are sufficient labeled instances. In many applications, such as object detection, document and web-page categorization, labeled instances however are difficult, expensive, or time consuming to obtain because they require empirical research or experienced human annotators. Semi-supervised learning algorithms use not only the labeled data but a...
متن کاملWeb Page Categorization using Multilayer Perceptron with Reduced Features
The web is a huge repository of knowledge and numerous hyperlinks. Web also serves a broad diversity of user communities and global information service centers. Every day the knowledge in web page upwards rapidly. Web pages can be used to convey the knowledge to web users. Such voluminous size of the web makes an intricacy of web information retrieval, web content filtering and web structure mi...
متن کامل