- From: Karl Dubost <karl+w3c@la-grange.net>
- Date: Sun, 17 Oct 2010 22:51:54 -0400
- To: public-privacy@w3.org
The research articles are usually interesting because they tell you a lot about what comes next. I found this article on the always interesting Microsoft Research site. How to Tell an Airport from a Home: Techniques and Applications Andreas Pitsillidis, Yinglian Xie, Fang Yu, Martin Abadi, Geofferey M. Voelker, and Stefan Savage October 2010 http://research.microsoft.com/apps/pubs/default.aspx?id=139079 5. SUMMARY This paper explores the notion of “location context”. Specifically, we develop techniques to derive context about the locations where users access the network (home, work, travel) by analyzing user mobility patterns from large service logs. We show that these techniques can scale to the Internet despite a number of practical challenges. Using concrete examples, we show that “location context” can provide interesting characteristics of users and their activities for a variety of applications. In the PDF, you can read that in section 4: The classification of home and travel IP addresses pro- vides new opportunities for online applications to under- stand user profiles. Based on the location categories of a user, we could infer their interests or intentions to customize search results and to target advertisement. For example, travel-related links may have higher rankings when the query comes from a travel IP address. -- Karl Dubost Montréal, QC, Canada http://www.la-grange.net/karl/
Received on Monday, 18 October 2010 02:52:09 UTC