Put out a few topics for discussion:

1
Introduction

• Put out a few topics for discussion
• Scope of weblog spam
• Incentives to spam
• Personalization and spam

2
Most blogs have no regular readers
Bloglines was the dominant feed reader with 33
market share in January 2005 according to Dick
Costolo, Feedburner, http//blogs.feedburner.com/f
eedburner/archives/000961.html Total weblogs from
Dave Sifry, Technorati, http//www.technorati.com/
weblog/2005/10/53.html Bloglines subscription
metrics from Jim Lanzone, Ask.com,
http//blog.ask.com/2005/10/deep_thoughts_a.html
3
Scale attracts spam

• Web started without spam
• As audience grew, profit motive grew
• Weblogs started without spam
• As audience grew, profit motive grew
• Hitting the mainstream attracts spam

4
Winner takes all attracts spam

• Getting the top slot Jackpot!
• Everyone will see it
• Personalization as a solution?
• Each person sees different content
• Targets differing perceptions of relevance

5
Findory.com

• Recommender system News and weblogs
• Learns from clickstream
• Every reader sees a different page
• Reduces winner takes all
• Crawl focuses on weblogs with readers
• Reduces spam blogs

6
Greg Linden

• Findory.com
• Amazon.com, Personalization
• University of Washington CS in AI
• Stanford Business School