期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2015
卷号：73
期号：2
出版社：Journal of Theoretical and Applied
摘要：User defined keyword search have limitation in their results and deliverance based on first indications. They face a challenge against social media analysis. Searches based on user defined keyword results in poor quality and consumes more time. All relevant and irrelevant information present in a link becomes difficult for the users to extract. So, ontology is used for effective monitoring analysis. The main challenge is to create domain-specific ontology in social media analysis. The earlier heuristic algorithm, HITS(Hyperlink Induced Topic Search) was proposed for analyzing authority and hubs of pages, but since each site had extra hyperlinks like navigation panels, HITS failed in providing good precision.To overcome this difficulty, LAMIS(Link Analysis on Mining Web Informative Structures). LAMIS uses information entropy to give a higher precision from 133% to 232% and recall gets improved between 0.5 and 1 while retrieving the documents form online social media.
