Sobre publicidade direcionada baseada em conteúdo

• Ano de defesa: 2006
• Autor(a) principal: Marco Antonio Pinheiro de Cristo
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: Publicidade na internet; Recuperação de informação; World Wide Web (Sistema de recuperação da informação); World Wide Web Publicações; Computação; Editoração eletronica; Editoração da Web
• Link de acesso: http://hdl.handle.net/1843/RVMR-6TCNUJ

The current boom of online advertising is associated with the revenuesoriginated from search advertising, which has become the driving force sustaining monetization of Web services. According to Forrester Research, search advertising revenues were projected to grow from US $3.6 billion in 2004 to US $11.6 billion by 2010. Actually, numbers might be quite larger. To illustrate, Yahoo reported search advertising revenues in the total amount of US $875 million for the second quarter of 2005 only, while Google reported revenues in the total amount of US $1.384 billion for the same period. Further, forecasts suggest that the influence of search advertising will increase in the upcoming years through diversification and the production of new types of search-related advertising. This rapidly consolidating market involves complex business networks and increasingly sophisticated technology. Thus, the exploitation of new forms of search advertising requires advancesin the commercial and in the technology front. In this work, we discuss the use of Information Retrieval (IR) techniques to improve the performance of ad placement methods in search advertising, with emphasis on content-targeted advertising. We investigate how the different sources of evidence already available to information gatekeepers (that operate keyword-targeted advertising systems) affect the matching of ads tothe content of aWeb page. As a result of this analysis, we propose new strategies for associating advertisements with Web pages. Experiments with a real ad collection show that the proper use of the available sources of evidence can lead to high quality matching algorithms.We also exploit the combination of conceptual and syntactical evidence.To accomplish this, we first study how to improve Web document classification. We observe that methods based on link evidence can be successfully used to improve the classification based only on content. We then use the best classifiers obtained as source of conceptual information and conclude that matching algorithms in content-targeted advertising can be enhanced by combining the context-based ranking obtained through manual and automatic classification with the ranking provided by the syntactical matching methods.