Applying effective feature selection techniques with hierarchical mixtures of experts for spam classification

Belsis, P; Fragos, K; Gritzalis, S; Skourlas, C

dc.contributor.author	Belsis, P	en
dc.contributor.author	Fragos, K	en
dc.contributor.author	Gritzalis, S	en
dc.contributor.author	Skourlas, C	en
dc.date.accessioned	2014-03-01T01:27:56Z
dc.date.available	2014-03-01T01:27:56Z
dc.date.issued	2008	en
dc.identifier.issn	0926227X	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/18648
dc.subject	Hierarchical mixtures of experts	en
dc.subject	Machine learning based processing	en
dc.subject	Spam mail	en
dc.subject.other	Cost effectiveness	en
dc.subject.other	Data storage equipment	en
dc.subject.other	Electronic mail	en
dc.subject.other	Internet service providers	en
dc.subject.other	Learning systems	en
dc.subject.other	Mixtures	en
dc.subject.other	Random processes	en
dc.subject.other	Robot learning	en
dc.subject.other	Spamming	en
dc.subject.other	Cost efficiencies	en
dc.subject.other	Data dimensionalities	en
dc.subject.other	Data samples	en
dc.subject.other	Effective algorithms	en
dc.subject.other	Feature selections	en
dc.subject.other	Hierarchical mixtures	en
dc.subject.other	Hierarchical mixtures of experts	en
dc.subject.other	Internet services	en
dc.subject.other	Linear relationships	en
dc.subject.other	Machine learning based processing	en
dc.subject.other	Machine learnings	en
dc.subject.other	Massive quantities	en
dc.subject.other	Non linearities	en
dc.subject.other	Offensive languages	en
dc.subject.other	Perceptron	en
dc.subject.other	Spam classifications	en
dc.subject.other	Spam filtering	en
dc.subject.other	Spam mail	en
dc.subject.other	Statistical models	en
dc.subject.other	Storage resources	en
dc.subject.other	Text filtering	en
dc.subject.other	Unsolicited bulk e-mails	en
dc.subject.other	Internet	en
dc.title	Applying effective feature selection techniques with hierarchical mixtures of experts for spam classification	en
heal.type	journalArticle	en
heal.identifier.primary	10.3233/JCS-2008-0319	en
heal.identifier.secondary	http://dx.doi.org/10.3233/JCS-2008-0319	en
heal.publicationDate	2008	en
heal.abstract	E-mail abuse has been steadily increasing during the last decade. E-mail users find themselves targeted by massive quantities of unsolicited bulk e-mail, which often contains offensive language or has fraudulent intentions. Internet Service Providers (ISPs) on the other hand, have to face a considerable system overloading as the incoming mail consumes network and storage resources. Among the plethora of solutions, the most prominent in terms of cost efficiency and complexity are the text filtering approaches. Most of the approaches model the problem using linear statistical models. Despite their popularity - due both to their simplicity and relative ease of interpretation - the non-linearity assumption of data samples is inappropriate in practice. This is mainly due to the inability of other approaches to capture the apparent non-linear relationships, which characterize these samples. In this paper, we propose a margin-based feature selection approach integrated with a Hierarchical Mixtures of Experts (HME) system, which attempts to overcome limitations common to other machine-learning based approaches. By reducing the data dimensionality using effective algorithms for feature selection we evaluated our system with publicly available corpora of e-mails, characterized by very high similarity between legitimate and bulk e-mail (and thus low discriminative potential). We experimented with two different architectures, a hierarchical HME and a perceptron HME. As a result, we confirm the domination of our Spam Filtering (SF) - HME method against other machine learning approaches, which present lesser degree of recall, as well as against traditional rule-based approaches, which lack considerably in the achieved degrees of precision. © 2008 - IOS Press and the authors. All rights reserved.	en
heal.journalName	Journal of Computer Security	en
dc.identifier.doi	10.3233/JCS-2008-0319	en
dc.identifier.volume	16	en
dc.identifier.issue	6	en
dc.identifier.spage	761	en
dc.identifier.epage	790	en