Ranking techniques for finding correlated webpages

Abstract

In general,when users try to search information, they can have difficulties to express the information as exact queries. Therefore, users consume many times to find useful webpages. Previous techniques could not solve the problem effectively. In this paper, we propose an algorithm, RCW (Ranking technique for finding Correlated Webpages) for improving previous ranking techniques. Our method makes it possible to retrieve not only basic webpages but also correlated webpages. Therefore, RCW algorithm in this paper can help users easily look for meaningful information without using exact queries. To find correlated webpages, the algorithm applies a novel technique for computing correlations among webpages. In performance evaluation, we test precision, recall, and NDCG of our RCW compared with the other popular system. In this result, RCW guarantees that itfinds the number of correlated webpages greater than the other method, and shows high ratios in terms of precision, recall, and NDCG. © 2013 Springer Science+Business Media.