id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_kaayfpvw3rcxtdhrfkbkaaatpi	Wen-Feng Hsiao	An incremental cluster-based approach to spam filtering	2008	10	.pdf	application/pdf	6621	686	64	To filter spam from legitimate emails, automatic classification approaches using text mining In the first phase, it clusters emails in each given class into several groups, and an equal number of features (keywords) are extracted from each group to manifest the features in the minority class. of skewed and changing class distributions, and its incremental learning can also reduce the cost of re-training. Keywords: Email classification; Skewed class distribution; Concept drift; Incremental learning employed simple and straightforward methods, called filtering by rules, which classify spam emails by matching skewed class distribution and of no concept drift, we examine ICBC's performance using the whole collected data Experiment I To examine the performance of ICBC under the condition of no significant class-skewed problem and of no concept drift To deliberately make significant skewed class distributions, we select five training data sets, each of which consists of 200 spam and 10, 20, 30, 40, or 50 legitimate	./cache/work_kaayfpvw3rcxtdhrfkbkaaatpi.pdf	./txt/work_kaayfpvw3rcxtdhrfkbkaaatpi.txt