Abstract
Survey articles provide a comprehensive overview of a specific area of research. Automatic detection of survey articles from huge scientific literature is interesting and useful knowledge discovery task in academic social networks. There are different features which can be exploited to differentiate between survey articles and other research articles. Surveys articles are usually citing many important articles this important feature is used in the past for finding surveys using HITS algorithm in addition to base words, base cues, and article length features. The rank of authors writing the articles and text of articles is not considered. In this paper, two additional features based on Author Rank (author authority score of her papers) and textual feature Entropy (paper disorder score) are introduced. Entropy feature has its special significance as it can be used even when there is no link structure. Empirical results show that proposed enhancements are useful and better results are obtained. Especially for large number of top n papers our proposed methods performance is very stable as compared to existing methods.