User-aware advertisability

Hai Tao Yu, Tetsuya Sakai

研究成果

抄録

In sponsored search, many studies focus on finding the most relevant advertisements (ads) and their optimal ranking for a submitted query. Determining whether it is suitable to show ads has received less attention. In this paper, we introduce the concept of user-aware advertisability, which refers to the probability of ad-click on sponsored ads when a specific user submits a query. When computing the advertisability for a given query-user pair, we first classify the clicked web pages based on a pre-defined category hierarchy and use the aggregated topical categories of clicked web pages to represent user preference. Taking user preference into account, we then compute the ad-click probability for this query-user pair. Compared with existing methods, the experimental results show that user preference is of great value for generating user-specific advertisability. In particular, our approach that computes advertisability per query-user pair outperforms the two state-of-the-art methods that compute advertisability per query in terms of a variant of the normalized Discounted Cumulative Gain metric.

本文言語English
ホスト出版物のタイトルInformation Retrieval Technology - 9th Asia Information Retrieval Societies Conference, AIRS 2013, Proceedings
ページ452-463
ページ数12
DOI
出版ステータスPublished - 2013
イベント9th Asia Information Retrieval Societies Conference on Information Retrieval Technology, AIRS 2013 - Singapore, Singapore
継続期間: 2013 12 92013 12 11

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
8281 LNCS
ISSN(印刷版)0302-9743
ISSN(電子版)1611-3349

Conference

Conference9th Asia Information Retrieval Societies Conference on Information Retrieval Technology, AIRS 2013
国/地域Singapore
CitySingapore
Period13/12/913/12/11

ASJC Scopus subject areas

  • 理論的コンピュータサイエンス
  • コンピュータ サイエンス(全般)

フィンガープリント

「User-aware advertisability」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル