Position-wise contextual advertising: Placing relevant ads at appropriate positions of a web page Zongda Wu a,b , Guandong Xu c , Chenglang Lu a,d,n , Enhong Chen b , Yanchun Zhang e , Hong Zhang f a Wenzhou University, Wenzhou 325035, Zhejiang, China b University of Science and Technology of China, Hefei 230026, Anhui, China c University of Technology, Sydney, Australia d Northwestern Polytechnical University, Xi'an 710072, China e Victoria University, Melbourne, Australia f Institute of Science and Technology Information of Zhejiang Province, Hangzhou 310006, China article info Article history: Received 4 July 2012 Received in revised form 7 April 2013 Accepted 7 April 2013 Communicated by Tao Mei Available online 15 June 2013 Keywords: Wikipedia knowledge Similarity Contextual advertising abstract Web advertising, a form of online advertising, which uses the Internet as a medium to post product or service information and attract customers, has become one of the most important marketing channels. As one prevalent type of web advertising, contextual advertising refers to the placement of the most relevant ads at appropriate positions of a web page, so as to provide a better user experience and increase the user's ad-click rate. However, most existing contextual advertising techniques only take into account how to select as relevant ads for a given page as possible, without considering the positional effect of the ad placement on the page, resulting in an unsatisfactory performance in ad local context relevance. In this paper, we address the novel problem of position-wise contextual advertising, i.e., how to select and place relevant ads properly for a target web page. In our proposed approach, the relevant ads are selected based on not only global context relevance but also local context relevance, so that the embedded ads yield contextual relevance to both the whole target page and the insertion positions where the ads are placed. In addition, to improve the accuracy of global and local context relevance measure, the rich wikipedia knowledge is used to enhance the semantic feature representation of pages and ad candidates. Last, we evaluate our approach using a set of ads and pages downloaded from the Internet, and demonstrate the effectiveness of our approach. & 2013 Elsevier B.V. All rights reserved. 1. Introduction Web advertising is becoming an increasingly important and popular advertising market today. PwC 1 predicts that web advertis- ing will become the second largest advertising medium in America after TV within the next four years, and spending in this area will increase from 24 billion dollars in 2009 to 34 billion dollars in 2014. A large part of web advertising consists of textual ads, which are short text messages usually marked as “sponsored links” or similar. Now, there are two main types of textual web advertising, i.e., sponsored search and contextual advertising [1,2]: 1. Sponsored search (also called keyword-targeted advertising), which selects ads based on keywords contained in search queries given by users, is characterized by placing paid textual ads links on the result pages returned by a web search engine (e.g., Google). 2. Contextual advertising (also called content-targeted advertising), which judges the context relevance of ads to the page that the user is browsing, refers to the selection of relevant commercial ads for the target page. One of the important advantages of contextual advertising over the sponsored search is that it can support various types of web sites, which range from individual bloggers and small niche communities to large publishers (e.g., major newspapers). Now, almost all for-profit non-transactional sites, i.e., the sites that do not sell anything directly, rely heavily on the revenues from contextual advertising. Without contextual ads, the Web will lose the most of its market value. The first major contextual advertising platform was provided by Google in 2003 [3]. Now, almost all popular search engines such as Baidu, Yahoo! and Microsoft Bing provide similar platforms for ad Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/neucom Neurocomputing 0925-2312/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.neucom.2013.04.018 n Corresponding author at: Wenzhou University, Wenzhou 325035, China. E-mail addresses: zongda1983@163.com (Z. Wu), guandong.xu@vu.edu. au (G. Xu), chenglang.lu@qq.com (C. Lu), cheneh@ustc.edu.cn (E. Chen). 1 PricewaterhouseCoopers, a global professional services company —http://www.pwc.com. Neurocomputing 120 (2013) 524–535