Development of web spam Research

This year's workshop on Web spam has been suspended. Does this mean that there is no research room for this web Spam? The answer is no. According to the research on the latest data set clueweb09 dataset, For the first 10 results of some popular queries, most pages (especially English pages) are cheating pages [1]. This shows that cheating is still very popular and has serious consequences for information retrieval. This also shows from some perspective that our current research on anti-cheating strategies is not very successful, and there is still a long way to go. Therefore, there is still much room for research on Web spam. However, a simple method may have been developed, so it is very challenging to study web spam.

I consulted the organizer about why airweb was stopped. The other party's explanation is that airweb is already running very well. If you want to continue, it will become a conference, but he is already a co-chair of WWW 2011, I don't want to hold another meeting (for example, WWW and wsdm), so I stopped airweb. In addition, he also said that in 2011, a new workshop about web quality may be opened. However, there are not many chanllenge participants on this topic in ecml/pkdd, so there are still a lot of uncertainties in the future.

Personally, I think there are many other aspects worth studying in the Web spam, but there may be some bottlenecks. As Web 2.0 gradually leaves people's lives, social web sites is becoming more and more popular. The resulting cheating problem is called social spam. The Research on anti social spam is currently in its infancy. We are looking for several attributes and then building a classifier to solve this problem using machine learning methods. The research in this direction is worth exploring.

[1]. Cormack, G. V., M.D. Smucker, and C. Clarke, efficient and valid tive spam filtering and re-ranking for large Web datasets. 2010.