Jobs

Clickfraud is a significant concern within the online advertising industry. The media is making breathtaking claims about the extent and scope of the problem, with some media sources estimating that the problem is as big as $1 billion per year. Companies such as Yahoo are now facing law suits because their syndicated links have been used by spyware programs, and Google has already settled over $90 million dollars.

A potential technique used by fraudsters is to compromise a large network of computers, known as a bot-net, and use this distributed network to click on pay per click ads so as to generate fraudulent revenue stream. Botnets are a serious problem not only for the unfortunate persons who's computers have been compromised, but also for advertisers who may be the victim of a botnet attack. Microsoft is working on methods to address the problem in all of its aspects, and make the internet safe for advertisers.

The Clickfraud Data Mining Scientist is the primary line of defense against botnet attacks against adCenter. This candidate will monitor activity on a day-to-day basis, analyze and understand new types of attacks, develop countermeasures, and then make modifications to adCenter's machine learning algorithms to catch and neutralize these threats. This position requires an ability to think like a fraudster, understand money-making schemes, and identify ways of closing down these holes.

The candidate will be responsible for adCenter revenue and detection rate performance, and will be able to present on new trends and threats in this area.

The successful candidate must have a range of skills, be able to get down into the details and be hands on, and drive the development of counter-measures. Skills include:

(a) Machine learning and Data Mining experience: candidate should be able to implement updates to Smartpricing, Realtime and Offline filtration models.

(b) Syndication team support: Candidate will analyze new syndication publishers before they come on-line and determine the quality of their traffic sources.

(a) Database experience: Experience with large-scale database deployments is desired. Microsoft uses SQLServer.

(b) Case Base Development: Identify cases of robotic activity and set these aside for training data.

(c) Statistical testing: Individual should be able to apply rigorous statistical methods to web usage data, in order to determine which behavioral signatures are statistically unusual.

(d) Passion and commitment: Individual should be passionate about this area and willing to commit their skills and abilities to this area.

(e) Collaboration with Crimes Unit and Law Enforcement: Individual will have the opportunity to collaborate with ex-Scotland Yard and Department Of Justice officials working in Microsoft as part of our Internet Crimes Task Force. Bot nets discovered will be referred to law enforcement who will pursue legal remedies.

(f) Results driven personality: Individual will be responsible for day-to-day revenue performance and detection performance of adCenter. Individual should

(a) Communications: Individual should possess superb written and oral communications skills. Individual should be able to communicate to both executives - in explaining revenue numbers - and developers - in explaining new countermeasures that are needed.

Contact adCenter today and ask about Clickfraud Data Mining and let's work together to combat this major new problem.

Contact:
In order to be considered for this position, you must apply online via our career site: Clickfraud Data Mining Scientist, Requisition / Job Code: 204514.

If you have questions about this position or you would like to confirm that I have received your application, please contact me at: lasham@microsoft.com. Resumes received via email cannot be considered until you formally apply online.