PEBL: Web page classification without negative examples

By Hwanjo Yu, Jiawei Han, Kevin Chen-Chuan Chang. IEEE Transactions on Knowledge and Data Engineering, 2004

301 citations recorded by the source corpus.

Web page classification is one of the essential techniques for Web mining because classifying Web pages of an interesting class is often the first step of mining the Web. However, constructing a classifier for an interesting class requires laborious preprocessing such as collecting positive and negative training examples. For instance, in order to construct a "homepage" classifier, one needs to collect a sample of homepages (positive examples) and a sample of nonhomepages (negative examples). In particular, collecting negative training examples requires arduous work and caution to avoid bias. The paper presents a framework, called positive example based learning (PEBL), for Web page classification which eliminates the need for manually collecting negative training examples in preprocessing. The PEBL framework applies an algorithm, called mapping-convergence (M-C), to achieve high classification accuracy (with positive and unlabeled data) as high as that of a traditional SVM (with positive and negative data). M-C runs in two stages: the mapping stage and convergence stage. In the mapping stage, the algorithm uses a weak classifier that draws an initial approximation of "strong" negative data. Based on the initial approximation, the convergence stage iteratively runs an internal classifier (e.g., SVM) which maximizes margins to progressively improve the approximation of negative data. Thus, the class boundary eventually converges to the true boundary of the positive class in the feature space. We present the M-C algorithm with supporting theoretical and experimental justifications. Our experiments show that, given the same set of positive examples; the M-C algorithm outperforms one-class SVMs, and it is almost as accurate as the traditional SVMs.

Subfields: Artificial Intelligence, Information Systems, Molecular Biology

View Paper: https://www.semanticscholar.org/paper/5bad19f3a4ad72956de997b3e5bf887901300c65

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Information Systems

In Molecular Biology

Citations in the corpus

Learning classifiers from only positive and unlabeled data (2008, methodology): “Papers using this general approach include [24, 20, 23], and the idea has been rediscovered independently a few times, most recently in [22, Section 2.”
Learning from positive and unlabeled data: a survey (2018, methodology): “S-EM [63] Spy EM NB ∆E Roc-SVM [54] Rocchio Iterative SVM FNR > 5% Roc-Clu-SVM [54] Rocchio∗ Iterative SVM FNR > 5% PEBL [114,113] 1-DNF Iterative SVM Last A-EM [56] Augmented Negatives EM NB ∆F LGN [55] Single Negative BN / PE PUC [115] PE (EM) NB Unspecified WVC/PSOC[81] 1-DNF∗ Iterative SVM Vote ”
Introduction to the special topic section on mining Web resources for enhancing information retrieval (2007, methodology): “Techniques such as PEBL learning framework ( Yu et al., 2004 ), semisupervised learning (Nigam, McCallum, Thrun, & Mitchell, 2000), singleclass learning (Denis, 1998), and one class support vector machine (OSVM) (Manevitz & Yousef, 2002) have been adopted to solve the problem.”
A Classification Framework for Anomaly Detection (2005, methodology): “A well-known heuristic (see e.g. Fan et al., 2001; González and Dagupta, 2003; Yu et al., 2004; Theiler and Cai., 2003) for anomaly detection is to generate a labeled data set by assigning one label to the original unlabeled data and another label to a set of artificially generated data, and then ap”
A Survey of Recent Trends in One Class Classification (2009, methodology): “PEBL is sensitive to the number of positive examples and gives poor results when they are small in number.”
One-class classification: taxonomy of study and review of techniques (2013, unknown): “Their experiments are performed on the Reuters data set and compared against PEBL and OSVM and it is shown that the GA based classification performs better.”
Outlier Analysis (2013, background): “For very large scale collections such as the web and social networks [493], this scenario is quite common.”
New classification models for detecting Hate and Violence web content (2015, background): “…link structure (Chakrabarti et al, 1998, Cohen, 2002), hierarchical structure (Dumais and Chen, 2000), web page classification with Positive Example Based Learning (PEBL) (Yu et al, 2004) and probabilistic relational models (Getoor et al, 2001; Broecheler et al, 2010; Fersini and Messina, 2013).”