Abstract:In this chapter, the classification performance of latent semantic indexing (LSI) applied to the task of detecting and filtering unsolicited bulk or commercial email (UBE, UCE, commonly called "spam") is studied. Comparisons to the simple vector space model (VSM) and to the extremely widespread, de-facto standard for spam filtering, the SpamAssassin system, are summarized. It is shown that VSM and LSI achieve significantly better classification results than SpamAssassin. Obviously, the classification performance achieved in this special application context strongly depends on the feature sets used. Consequently, the various classification methods are also compared using two different feature sets: (1) a set of purely textual features of email messages that are based on standard word- and token-extraction techniques, and (2) a set of application-specific "meta features" of email messages as extracted by the SpamAssassin system. It is illustrated that the latter tends to achieve consistently better classification results. A third central aspect discussed in this chapter is the issue of problem reduction in order to reduce the computational effort for classification, which is of particular importance in the context of time-critical on-line spam filtering. In particular, the effects of truncation of the SVD in LSI and of a reduction of the underlying feature set are investigated and compared. It is shown that a surprisingly large amount of problem reduction is often possible in the context of spam filtering without heavy loss in classification performance.

Application of LS-SVM in Spam-filtering

An Imbalanced Spam Mail Filtering Method

Research on Spam Filtering Technology Using Support Vector Machine

A Chinese Anti-Spam Filter Approach Based on Support Vector Machine

A Spam Filter Approach with the Improved Machine Learning Technology

Study on the Spam-Filtering System based on Feature Selection Mechanism and Improved SVM Classification

SVM-Based Spam Filter with Active and Online Learning.

Combining Svm With Orthogonal Centroid Feature Selection For Spam Filtering

Spam Filtering Based on Latent Semantic Indexing

Novel method for Chinese spam detection based on one-class support vector machines

Classify E-mails by Support Vector Machine

Filtering Chinese Spam Email Using Logistic Regression

Active Learning for Online Spam Filtering

An Improved E-mail Classifier Based on Support Vector Machine

Image Spam Filtering Based on Gradient and Color Feature

SVM Classification:Its Contents and Challenges

A Fast Content-Based Spam Filtering Algorithm with Fuzzy-SVM and K-means

Spam Filter Algorithm with Improved Porter Stemmer and Kernels Methods

Interactive Support Vector Machine Learning Algorithm and Its Application

Spam image discrimination using support vector machine based on higher-order local autocorrelation feature extraction

Spam Message Self-Adaptive Filtering System Based on Naive Bayes and Support Vector Machine