Proceedings of the 11th International Conference on Electronic Commerce
FindWDO: a k-nearest neighbors approach for detecting Web document outliers
ACST '08 Proceedings of the Fourth IASTED International Conference on Advances in Computer Science and Technology
Web content outlier mining through mathematical approach and trust rating
ACACOS'11 Proceedings of the 10th WSEAS international conference on Applied computer and applied computational science
International Journal of Computational Science and Engineering
Hi-index | 0.00 |
Outlier mining is dedicated to finding data objects which differ significantly from the rest of the data. Outlier mining has been extensively studied in statistics and recently data mining. However, exploring the web for outliers has received very little attention in the mining community. Web content outliers are documents with ývarying contentsý compared to similar web documents taken from the same domain. Mining web content outliers may lead to the identification of competitors and emerging business patterns in electronic commerce. This paper proposes WCOND-Mine algorithm for mining web content outliers using n-grams without a domain dictionary. Experimental results with embedded motifs show that WCOND-Mine is capable of finding web content outliers from web datasets.