WWW 2008 / Workshop Summary April 21-25, 2008 ˇ Beijing, China Fourth International Workshop on Adversarial Information Retrieval on the Web (AIRWeb 2008) Carlos Castillo chato@yahoo-inc.com ABSTRACT Adversarial IR in general, and search engine spam, in particular, are engaging research topics with a real-world impact for Web users, advertisers and publishers. The AIRWeb workshop will bring researchers and practitioners in these areas together, to present and discuss state-of-the-art techniques as well as real-world exp eriences. Given the continued growth in search engine spam creation and detection efforts, we exp ect interest in this AIRWeb to surpass that of the previous three editions of the workshop (held jointly with WWW 2005, SIGIR 2006, and WWW 2007 resp ectively). Yahoo! Research Barcelona, Spain Kumar Chellapilla kumarc@microsoft.com Microsoft Live Labs Redmond, WA, USA fetterly@microsoft.com Microsoft Research Mountain View, CA, USA Dennis Fetterly Categories and Subject Descriptors H.3.5 [Information Storage and Retrieval]: Online Information Services General Terms Documentation Keywords Adversarial information retrieval, web spam, spamdexing, search engine spam growth and success. However, this op enness is also a ma jor source of new challenges for information retrieval methods. Search engine spam is not a new problem; it has b een an imp ortant issue for commercial providers for a numb er of years, and is not likely to b e solved in the near future. Web spam damages search engine reputation. It exploits and as a result weakens the trust relationship b etween users and search engines [6]. According to Henzinger et al. [7], "Spamming has b ecome so prevalent that every commercial search engine has had to take measures to identify and remove spam. Without such measures, the quality of the rankings suffers severely." On the "op en web" a naive application of ranking methods in no longer an option. For instance, PageRank [8] in its pure form is very susceptible to spam: the authors of [4] ranked 100 million pages using PageRank and found that 11 out of the top 20 were p ornographic and achieved such high ranking through link manipulation. Adversarial information retrieval is a research area in which several things remain to b e discovered. Sahami et al. [9] have noted that "Adversarial classification is an area in which precious little work has b een done, but effective methods can provide large gains." Also, adversarial IR problems can b e approached from many different p ersp ectives, including information retrieval, machine learning and game theory. 1. INTRODUCTION Before the advent of the World Wide Web, information retrieval algorithms were develop ed for relatively small and coherent document collections such as newspap er articles or b ook catalogs in a library. In comparison to these collections, the Web is massive, much less coherent, changes more rapidly, and is spread over geographically distributed computers [1]. Scaling information retrieval algorithms to the World Wide Web is a challenging task. Success to date is depicted by the ubiquitous use of search engines to access Internet content. From the p oint of view of a search engine, the Web is a mix of two typ es of content: the "closed Web" and the "op en Web" [2]. The closed web comprises a few high-quality controlled collections which a search engine can fully trust. The "op en Web," on the other hand, includes the vast ma jority of Web pages, which lack an authority asserting their quality. The op enness of the Web has b een the key to its rapid Copyright is held by the author/owner(s). WWW 2008, April 21­25, 2008, Beijing, China. ACM 978-1-60558-085-2/08/04. 2. WORKSHOP TOPICS Adversarial Information Retrieval addresses tasks such as gathering, indexing, filtering, retrieving and ranking information from collections wherein a subset has b een manipulated maliciously [5]. On the Web, the predominant form of such manipulation is "search engine spamming" or spamdexing, i.e.: malicious attempts to influence the outcome of ranking algorithms, aimed at getting an undeserved high ranking for some items in the collection. There is an economic incentive to rank higher in search engines, considering that a good ranking on them is strongly correlated with more traffic, which often translates to more revenue [10]. As in previous years, automatic detection of search engine spam is exp ected to b e the dominant theme of this workshop. Three basic forms of web spam are included: ˇ Link spam ˇ Content spam ˇ Cloaking 1267 WWW 2008 / Workshop Summary Several other adversarial IR topics that we welcome include: ˇ Blog spam filtering ˇ Click fraud detection ˇ Reverse engineering of ranking algorithms ˇ Web content filtering ˇ Advertisement blocking ˇ Stealth crawling ˇ ˇ ˇ ˇ ˇ ˇ ˇ ˇ ˇ ˇ ˇ ˇ April 21-25, 2008 ˇ Beijing, China Zolt´n Gyongyi ­ Stanford University a ¨ Monika Henzinger ­ Google Pranam Kolari ­ Yahoo! Applied Research Mark Manasse ­ Microsoft Research Marc Na jork ­ Microsoft Research Alexandros Ntoulas ­ Microsoft Research Jan Pedersen ­ Yahoo! Research Erik Selb erg ­ Amazon.com Torsten Suel ­ Polytechnic University Mike Thelwall ­ University of Wolverhampton Tao Yang ­ Ask.com Baoning Wu ­ Snap.com 3. WEB SPAM CHALLENGE In 2007, we introduced a novel element: the Web spam challenge. We released a reference collection for Web Spam Detection that comprises Web pages, a Web graph, and lab els for a subset of the pages. Web pages in this collection were lab eled as "normal" or "spam" by humans [3]. Using this data set, the challenge was to predict which pages in the unlab eled part of the data are spam and which are normal. For 2008, we released an up dated reference collection covering a significantly increased numb er of hosts. We also encouraged authors submitting pap ers on search engine spam to test their systems on the up dated reference collection. We ask that participating researchers submit predictions (normal/spam) for all unlab eled elements in the collection. Predictions will b e evaluated on a part of the collection for which human-provided lab els will b e held for testing. Results will b e announced at the AIRWeb 2008 workshop. The Web spam challenge serves a dual purp ose: it allows the comparison of different systems, which has not b een p ossible in the past for lack of a reference collection; and it stimulates research on this area given its comp etitive nature. 5. REFERENCES [1] Arvind Arasu, Junghoo Cho, Hector Garcia-Molina, Andreas Paep cke, and Sriram Raghavan. Searching the web. ACM Transactions on Internet Technology (TOIT), 1(1):2­43, 2001. [2] Terrence A. Brooks. Web search: how the Web has changed information retrieval. Information Research, 8(3), April 2003. [3] Carlos Castillo, Deb ora Donato, Luca Becchetti, Paolo Boldi, Stefano Leonardi, Massimo Santini, and Sebastiano Vigna. A reference collection for web spam. SIGIR Forum, 40(2):11­24, Decemb er 2006. [4] Nadav Eiron, Kevin S. Curley, and John A. Tomlin. Ranking the web frontier. In Proceedings of the 13th international conference on World Wide Web, pages 309­318, New York, NY, USA, 2004. ACM Press. [5] Dennis Fetterly. Adversarial information retrieval: The manipulation of web content. ACM Computing Reviews, July 2007. [6] Zolt´n Gy¨ngyi and Hector Garcia-Molina. Spam: It's a o not just for inb oxes anymore. IEEE Computer Magazine, 38(10):28­34, 2005. [7] Monika R. Henzinger, Ra jeev Motwani, and Craig Silverstein. Challenges in web search engines. SIGIR Forum, 36(2):11­22, 2002. [8] Lawrence Page, Sergey Brin, Ra jeev Motwani, and Terry Winograd. The PageRank citation ranking: bringing order to the Web. Technical rep ort, Stanford Digital Library Technologies Pro ject, 1998. [9] Mehran Sahami, Vibhu Mittal, Shumeet Baluja, and Henry Rowley. The happy searcher: Challenges in web information retrieval. In Trends in Artificial Intel ligence, 8th Pacific Rim International Conference on Artificial Intel ligence (PRICAI), pages 3­12, Auckland, New Zealand, August 2004. Springer. [10] Yi-Min Wang, Ming Ma, Yuan Niu, and Hao Chen. Spam double-funnel: connecting web spammers with advertisers. In WWW '07: Proceedings of the 16th international conference on World Wide Web, pages 291­300, New York, NY, USA, 2007. ACM Press. 4. WORKSHOP ORGANIZATION The proceedings of the workshop will b e published online in the ACM Digital Library, as well as distributed at the workshop. The workshop program has not b een finalized at the time of this writing. Once finalized, the program will b e available from the AIRWeb website: http://airweb.cse.lehigh.edu/2008/ 4.1 Program Committee We appreciate the service of the following researchers as Program Committee memb ers of the workshop: ˇ ˇ ˇ ˇ ˇ ˇ ˇ ˇ ˇ ˇ ˇ ˇ ˇ Einat Amitay ­ IBM Andr´s Benczur ­ Hungarian Academy of Sciences a ´ James Caverlee ­ Texas A&M University Paul-Alexandru Chirita ­ Adob e Gordon Cormack ­ University of Waterloo Nick Craswell ­ Microsoft Research Matt Cutts ­ Google Brian Davison ­ Lehigh University Ludovic Denoyer ­ University Paris 6 Aaron D'Souza ­ Google Edel Garcia ­ Mi Islita.com Natalie Glance ­ Nielsen BuzzMetrics Antonio Gulli ­ Ask.com 1268