Download PDFOpen PDF in browser

Application of Adaptive Neural Networks for the Filtration of Spam

9 pagesPublished: December 18, 2015

Abstract

The application of e-mail is the most widespread communication method. The increased email number includes permanently increased amount of unwanted notifications (spam). The statistics claims that 80% of the traffic is spam that lags the internet traffic, consumes space on hard disc, and reduces the capacity of the network, not to mention time spent for email’s classification. The advanced E-mail service packs contain spam filtration program. Nevertheless, to classify the e-mails as spam fully depends on the consumers, as information acceptable (or vital) for one consumer, may be a spam for other. Hence, it is crucial to have automatic spam filtration system for every individual consumer.
The paper discuss the solution, a learning system that filters emails through consumers’ parameters and gradually learn how to filter them accordingly. The system based on multi-layer neural network, which uses logistic activation function, learns via tutor and counter-spread algorithm of mistakes

Keyphrases: adaptive neural networks, neural networks, Spam, Spam Detection, spam filtration system

In: Georg Gottlob, Geoff Sutcliffe and Andrei Voronkov (editors). GCAI 2015. Global Conference on Artificial Intelligence, vol 36, pages 42--50

Links:
BibTeX entry
@inproceedings{GCAI2015:Application_of_Adaptive_Neural,
  author    = {Gela Besiashvili and Tamar Bliadze and Zurab Kochladze},
  title     = {Application of Adaptive Neural Networks for the Filtration of Spam},
  booktitle = {GCAI 2015. Global Conference on Artificial Intelligence},
  editor    = {Georg Gottlob and Geoff Sutcliffe and Andrei Voronkov},
  series    = {EPiC Series in Computing},
  volume    = {36},
  pages     = {42--50},
  year      = {2015},
  publisher = {EasyChair},
  bibsource = {EasyChair, https://easychair.org},
  issn      = {2398-7340},
  url       = {https://easychair.org/publications/paper/GtRd},
  doi       = {10.29007/t3rl}}
Download PDFOpen PDF in browser