Elsevier

Decision Support Systems

Volume 47, Issue 4, November 2009, Pages 364-373
Decision Support Systems

Ontological analysis of web surf history to maximize the click-through probability of web advertisements

https://doi.org/10.1016/j.dss.2009.04.001Get rights and content

Abstract

Due to an enormous influx of capital over the past decade, the online advertising industry has become extremely robust and competitive. The difference between success and failure in such a competitive market often rests in the ability to deliver advertisements that are closely in line with a user's interests. In this work, we propose and test a new online advertisement targeting technique which adapts and utilizes several powerful and well tested information retrieval and lexical techniques to develop an estimate of a user's affinity for particular products and services based on an analysis of a user's web surfing behavior. This new online ad targeting technique performs extremely well in our empirical tests.

Introduction

It is estimated that online advertisement revenues for the US alone will grow to $18.9 billion by 2010 [29]. Motivated by this upward trend in Internet advertising demand, many companies (e.g., Google, Yahoo, AOL, WSJ, etc.) have adopted a business model which is heavily dependent upon the revenue stream generated from their online advertisement publishing activities [10]. Although the business process has many similarities with ad publishing in traditional media settings, the online environment, primarily due to its transactional transparency, offers some very unique opportunities and challenges. This online transparency leads to an enormous amount of customer data which continues to drive the popularity and the importance of ad targeting [17]. According to Kessler and Acoido of the USA Today, “Microsoft is one of many companies collecting and aggregating data in new ways so sophisticated that many customers may not even realize that they're being watched.” It is acknowledged that a user's web surf history is and will be a very important component of this ad targeting process, but to date the research community has not provided a clear and dominant way to utilize this data. The extant literature has instead focused primarily on the utilization of demographic data. In this work, we propose and test an online ad targeting technique which is based upon the analysis of a user's web surfing habits. This technique shows great promise and can be used to complement other tools and in doing so offer an opportunity for online ad publishers to improve the efficiency and effectiveness of their online ad scheduling process, thereby enhancing their revenue generation ability.

The paper is organized as follows. Section 2 provides some background information with respect to the online advertising industry and discusses the relevant literature. Section 3 provides an introduction to information retrieval, structural representation and lexical reference systems. Section 4 introduces and discusses in detail the proposed Ontological Analysis of Web Surf History (OAWSH) Model. Section 5 provides the results of our model tests. And lastly, Section 6 provides conclusion and projected areas of future related research.

Section snippets

Background

There are three primary participants in the online advertising process. At the top of the chain is the Advertiser. This is a company that enters into an agreement with a publisher in order to enlist the publisher's assistance in the serving of their online advertisements. The associated ads are delivered to users of the publisher's Web pages. The Publisher is a company that expends resources in an effort to publish online advertisements in order to generate revenue. The Customer/User is the

Information retrieval introduction

Information retrieval (IR) is an area of research which attempts to extract usable information from textual data [11]. IR has historically been employed in the field of library sciences, but it has recently gained favor in many other fields including Internet search, cyber security and medicine [15]. The power of IR is its ability to handle textual information. IR has been applied in many domains, including document sorting, document retrieval, inference development and query response. We use

The ontological analysis of web surf history (OAWSH) model

The OAWSH model is based on an adapted VSM which incorporates lexical and structural information. To the best of our knowledge, this approach has never been considered for the online ad targeting problem. We are optimistic that the combination of these proven IR techniques will result in a powerful ad targeting model. We begin by giving a basic outline and process flow diagram of the model (refer to Fig. 2), and then provide a more thorough discussion of each step. In addition, for a very basic

Results

In this section, we report the results of the experimental analysis of the proposed model. Each user was asked to rank their level of interest on a scale of 1 to 5 for a set of ads, some of which were selected randomly and the remainder of which was selected based on one of the three weighting schemes. As discussed in Section 4.0, within the framework of our ad targeting process, we tested three different weighting schemes in an effort to identify the best html structural element weighting

Conclusion and future research

The purpose of this research is to introduce a new online ad targeting technique and to test its effectiveness in selecting a subset of targeted ads from a given advertisement corpus with respect to the overall level of interest to a given user base. The ultimate goal is to provide online advertisement publishers with another viable ad scheduling tool. The most commonly used method in industry is the random selection process and according to industry representatives, publishers are eagerly

Jason Deane is Assistant Professor of Business Information Technology in the Pamplin College of Business at Virginia Polytechnic Institute & State University. He received a Ph.D. in Decision and Information Sciences from the University of Florida, and an M.B.A. and B.S. in Business Administration from Virginia Tech. His current research interests are in the areas of artificial intelligence, computer aided decision support systems, information system security, large scale optimization and

References (35)

  • E. Agirre et al.

    Exploring automatic word sense disambiguation with decision lists and the web

  • S. Banerjee et al.

    An adapted lesk algorithm for word sense disambiguation using WordNet

  • A. Bhatnagar et al.

    Identifying locations for targeted advertisements on the internet

    International Journal of Electronic Commerce

    (2001)
  • P. Buiteclaar et al.

    Ranking and selecting synsets by domain relevance

  • F. Burkowski

    Retrieval activities in a database consisting of heterogeneous collections of structured text

  • M. Cecchini

    Quantifying the risk of financial events using kernel methods and information retrieval

  • P. Chatterjee et al.

    Modeling the clickstream: implications for web-based advertising efforts

    Marketing Science

    (2003)
  • D. Chickering et al.

    Targeted advertising on the web with inventory management

    Interfaces

    (2003)
  • A. Cucchiarelli et al.

    Automatic selection of class labels from a thesaurus for an effective semantic tagging of corpora

  • M. Dawande et al.

    Performance bounds of algorithms for scheduling advertisements on a web page

    Journal of Scheduling

    (2003)
  • W. Fan et al.

    On linear mixture of expert approaches to information retrieval

    Decision Support Systems

    (2006)
  • C. Fellbaum

    WordNet: An Electronic Lexical Database

    (1999)
  • J. Gonzalo et al.

    Indexing with WordNet synsets can improve text retrieval

  • F. Halasaz

    Reflections on notecards: seven issues for the next generation of hypermedia systems

  • A. Houston et al.

    Exploring the use of concept spaces to improve medical information retrieval

    Decision Support Systems

    (2000)
  • G. Karuga et al.

    AdPalette: an algorithm for customizing online advertisements on the fly

    Decision Support Systems

    (2001)
  • Kessler, M. and B. Acohido, Data Miners Dig a Little Deeper, USA Today...
  • Cited by (11)

    • Online banner advertisement scheduling for advertising effectiveness

      2020, Computers and Industrial Engineering
      Citation Excerpt :

      Purnamawati, Nababan, Tsani, Taqyuddin, and Rahmat (2018) proposed an advertisement scheduling problem with different advertisement frequencies between prime and regular times. Deane and Pathak (2009) stated the importance of online advertisement targeting, and Deane (2012) extended the problem of Deane and Agarwal (2012b) by considering advertisement targeting. Some studies proposed placement models and formulations in which two-dimensional display time slots are considered (Boskamp, Knoops, Frasincar, & Gabor, 2011; Anshu Gupta, Aggarwal, Kaul, & Jha, 2016; A Gupta, Manik, Aggarwal, & Jha, 2011; Kaul et al., 2018; Knoops, Boskamp, Wojciechowski, & Frasincar, 2009; Manik et al., 2016). da Silva, Schouery, and Pedrosa (2019)

    • Goal attainment on long tail web sites: An information foraging approach

      2013, Decision Support Systems
      Citation Excerpt :

      Patches (of the “non-stinky” variety) may also prove useful in online advertising. Recent research has examined the optimal design for online advertisements [17] and can be extended to include optimizing landing pages from learned patches as well. Another avenue of research may incorporate page content to form topically-relevant patches and trails, or even uncover intra-page patches.

    • Discovering Web Users’ Web Access Pattern Based on Psychology

      2020, Lecture Notes on Data Engineering and Communications Technologies
    • How does e-marketing influence consumer’s decision? A descriptive review

      2019, Journal of Advanced Research in Dynamical and Control Systems
    View all citing articles on Scopus

    Jason Deane is Assistant Professor of Business Information Technology in the Pamplin College of Business at Virginia Polytechnic Institute & State University. He received a Ph.D. in Decision and Information Sciences from the University of Florida, and an M.B.A. and B.S. in Business Administration from Virginia Tech. His current research interests are in the areas of artificial intelligence, computer aided decision support systems, information system security, large scale optimization and information retrieval.

    Dr. Praveen Pathak is an Assistant Professor of Decision and Information Sciences at the Warrington College of Business at the University of Florida. He received his PhD in Information Systems from the Ross School of Business, University of Michigan, Ann Arbor, in 2000. He also holds an MBA (PGDM) from the Indian Institute of Management, Calcutta, and an Engineering degree, B. Tech. (Hons.), from the Indian Institute of Technology, Kharagpur. His research interests include information retrieval, web mining, offshore outsourcing and business intelligence. His research has appeared in many journals such as Journal of Management Information Systems (JMIS), Decision Support Systems (DSS), IEEE Transactions on Knowledge and Data Engineering (TKDE), Information Processing and Management (IP&M), Journal of the American Society for Information Science and Technology (JASIST), and in leading information technology conferences such as ICIS, HICSS, WITS, and INFORMS.

    1

    Tel.: +1 352 392 9599.

    View full text