摘要
随着互联网数据的爆炸性增长,搜索引擎成为人们生活和工作中不可缺少的信息获取工具.作为搜索引擎盈利的主要商业模式之一,竞价排名的赞助商广告链接以其高效、低风险、灵活方便的特点受到传统营销方式下处于劣势的中小企业的青睐.然而搜索引擎结果页面上出现大量的广告链接是否影响用户体验?这些广告链接的实际收效如何?加入赞助商广告链接的搜索引擎如何影响用户的点击行为?开展面向这些问题的研究将对研究用户使用搜索引擎的行为特点、改进搜索引擎改进竞价排名的开展方式和指导中小企业如何更加有效地购买广告链接具有非常重要的指导意义.在总结国内外已经开展的相关研究的基础上,利用某商业搜索引擎提供的互联网用户点击日志和搜索引擎使用日志,采用数据挖掘和信息统计的方法,试图在全互联网的规模和商业搜索引擎的级别上提取用户和广告链接点击相关的信息,加以分析讨论,并对广告链接和非广告链接的相关性作深入分析.
With the explosive growth of information available on the Web,more and more users adopt search engines to collection information on the Internet.Meanwhile,sponsored search has become one of the most popular forms of Internet advertising because of its effectiveness and feasibility.As one of the predominant business model for Web search engines,sponsored search plays an important part in search engine studies.Sponsored search is popular among the small scale companies because it lowers risks and broadens market channels for these companies.But there remain many questions to be answered for sponsored search researches： Are sponsored results providing Web users good experience？Are Web search users disturbed by sponsored links on search engine result pages？How does sponsored search result affect the behavior of the search users？It is necessary to look into these questions and give advices to search engines and advertisers to make sponsored search more useful and acceptable.Click-through information is extracted from Web access logs collected by a commercial search engine in China.Based on analysis into this information,click through behavior on sponsored search results and ordinary results are compared.Experimental results show that some hypothesis proposed by previous researches don＇t accord with practical Web user behavior patterns.Meanwhile,a number of conclusions are also made to answer these proposed questions.
出处
《计算机研究与发展》
EI
CSCD
北大核心
2011年第1期133-138,共6页
Journal of Computer Research and Development
基金
国家自然科学基金项目(60736044
60903107)
高等学校博士学科点专项科研基金项目(20090002120005)
关键词
竞价排名
搜索引擎
用户日志挖掘
效果分析
互联网广告
sponsored search
search engine
user log mining
performance analysis
Web advertising
