作者: Baoning Wu
DOI:
关键词:
摘要: Web surfers rely on search engines to find information from the web. Search engine spam is attempt deceive ranking algorithms and considered by experts well-known companies be one of major challenges today. Without taking action, results will greatly harmed. This dissertation explores in detail effective solutions some techniques, such as link farms cloaking. Our approaches can effectively nullify farm effect our precision outperforms standard based algorithm, HITS, more than 200%. approach detecting cloaking behavior achieve an accuracy 90.5% a recall 86.8%. In addition, this studies idea combining topicality with trust demote spam. Experimental show that we up 43% TrustRank. We also investigate method authority improve quality. Experimental indicate significantly quality