Email overload: exploring personal information management of email
Proceedings of the SIGCHI Conference on Human Factors in Computing Systems
SIGIR '00 Proceedings of the 23rd annual international ACM SIGIR conference on Research and development in information retrieval
Automated email activity management: an unsupervised learning approach
Proceedings of the 10th international conference on Intelligent user interfaces
Adaptive anti-spam filtering for agglutinative languages: a special case for Turkish
Pattern Recognition Letters
MailRank: using ranking for spam detection
Proceedings of the 14th ACM international conference on Information and knowledge management
Spam Filtering based on Preference Ranking
CIT '05 Proceedings of the The Fifth International Conference on Computer and Information Technology
Automatic Personalized Spam Filtering through Significant Word Modeling
ICTAI '07 Proceedings of the 19th IEEE International Conference on Tools with Artificial Intelligence - Volume 02
Personalized Email Prioritization Based on Content and Social Network Analysis
IEEE Intelligent Systems
Enhanced email spam filtering through combining similarity graphs
Proceedings of the fourth ACM international conference on Web search and data mining
Support vector machines for spam categorization
IEEE Transactions on Neural Networks
Hi-index | 0.00 |
Email is one of the most successful computer applications in the Internet, and email-spam is also the biggest problem for users, preventing them to quickly process the important emails in a shortest time. In this paper, we propose an email recommender system using user actions and statistical methods. Instead of a two-class classification with Spam and Ham, we treat the problem as a multi-class classification in which each class is a recommendation action from user to an email. The most common actions are: reply, read and delete. An experiment is also conducted to test the framework, using Naïve Bayesian classifier and different threshold to evaluate the relations between number of features and the performance. The experiment shows a promising result with good prediction accuracy.