Colloquium Archive

Behavioral Advertising: The Offer You Can't Refuse

Chris Noofnagel, Boalt School of Law, University of California, Berkeley, CA


in 2009 and 2011, we surveyed top websites to determine how they were tracking consumers. We found that advertisers were using persistent tracking technologies that were relatively unknown to consumers. Two years later, we found that the number of tracking cookies expanded dramatrically and that advertisers had developed new, previously unobserved tracking mechanisms that users cannot avoid even with the strongest privacy settings. These empirical observations are valuable for the political debate surrounding online privacy, because the descriptive informs the framing and assumptions surrounding the merids of privacy law. In the political debate, "paternalism" is a frequently invoked objection to privacy rules. However, our empirical work demonstrates that advertisers use new, relatively unknown technologies to track people, specifically because consumers have no hear of these techniques. Futhermore, these technologies are used to obviate choice mechanisms that consumers exercise. Advertisers are a powerful force for the erosion of legal privacy protections, and increasingly, we demonstrate, they are also a froce that robs consumers of any technical ability to avoid online profiling. Our work inverts the assumption that privacy interventions are paternalistic while market approaches promote freedom. We empirically demonstrate that advertisers are behaving in highly paternalistic ways in order to keep consumers in the dark on privacy practices, and to make it impossible to avoid online tracking. Advertisers are so invested in the idea of a personalized web that they do not think consumers are competant to decide to reject it.

Autonomous Motion For Virtual Humans: Challenges And Possibilities

Michael Doherty, University of the Pacific, Stockton, CA


Creation of autonomous characters capable of human-like motion in virtual environments would appear to be achievable given available techniques in animation, artificial intelligence, and robotics. There are many examples of realistic motion in current applications, such as VR training simulations and video games. However, these current applications are limited in the sense that the range of possible behaviors is obviously restricted, resulting in strict limits on what the characters are capable of doing. In this talk, we'll examine the aspects of the problem that make arbitrary human-like motion difficult, and explore potential solutions that are currently being developed. From there we will attempt to extrapolate what will become possible as technology evolves and research continues.

How Safe Do Programming Languages Need To Be?

Kenneth Louden, San Jose State University, San Jose, CA


Throughout the history of programming language design there has been friction between the "freedom" and "discipline" camps: some prefer fewer programming rules, others want language-enforced compliance. In this talk, I will examine the notion of safety (or lack thereof) in programming languages and its relation to programming effort. I will also review some famous software disasters that indicate that "safe" languages rarely are as safe in practice as they are supposed to be. Indeed, the enforcement view appears to be waning, as perhaps it should. What should take its place? The answer should be obvious, with obvious but largely ignored consequences for computer science education.

Software For Self-Publishing Ebooks

Robert G. Plantz, Sonoma State University, Rohnert Park, CA


As ebooks become more popular we are learning that formatting for an ereader differs from a printed page, especially with science and technology books. I will discuss some of the issues and the free software for self-publishing an ebook. For example, LaTeX, docutils, docBook, Sigil, Sphinx.

Will Full Disk Encryption Keep My Data Safe?

Ron LaPedis, Seacliff Partners International, San Bruno, CA


There are many questions surrounding the topic of protecting sensitive information, and volume level and full disk encryption (FOE) seem to come up quite often as the answer. But are we asking the right questions? This colloquium will show you why FOE will not protect your data most of the time and will give you a tutorial of encryption technologies, validations, and injection points.