tion engine that minimizes RMSE make people happy?

 

Reader’s comment:

Another thing that seems to be often overlooked is how you get users to trust recommendations. When I first started playing with recommendation algorithms I was trying to produce novel results—things that the user didn’t know about and would be interesting to them, rather than using some of the more basic counting algorithms that are used for Amazon’s related products.

What I realized pretty quickly is that even I didn’t trust the recommendations. They seemed disconnected, even if upon clicking on them I’d realize they were, in fact, interesting and related.

What I came to from that was that in a set of recommendations you usually want to scale them such that you slip in a couple of obvious results to establish trust—things the user almost certainly knows of, and probably won’t click on, but they establish, “OK, yeah, these are my taste.” Then you

apply a second ranking scheme and jump to things they don’t know about. Once you’ve established trust of the recommendations they’re much more likely to follow up on the more novel ones.

—Scott Wheeler

 

from Jason hong’s “Privacy as… Sharing more information?”

What I am saying is that, rather than just viewing privacy as not sharing information with others, or viewing privacy as projecting a desired persona, we should also consider how to make systems so that people can safely share more information and get the associated benefits from doing so….

There are many dimensions here in this design space. We can change what is shared, how it is shared, when something is shared, and who it is shared with. One key challenge is in balancing privacy, utility, and the overhead for end users in setting up these policies. Another key challenge is understanding how to help people change these policies over time to adapt to people’s needs. These are issues I’ll discuss in future blog postings.

For me, a particularly intriguing way of thinking here is safe staging, an idea that Alma Whitten brought

to the attention of security specialists in her seminal paper Why Johnny Can’t Encrypt. The basic idea is that people progressively get more powerful tools as they become comfortable with a system, but are kept in a safe state as much as possible as they learn how to use the system. A real-world example would be training wheels on a bicycle. For systems that provide any level of awareness, the defaults might be set, for example, so that at first only close friends and family see anything, while over time people can easily share more information as they understand how the system works and how to control things.

 

from michael Stonebraker’s “DBmSs for Science applications: a Possible Solution”

Personally, I believe that there are a collection of planet-threaten-ing problems, such as climate change and ozone depletion, that only scientists are in a position to solve. Hence, the sorry state of DBMS support in particular (and system software support in general) for this class of users is very troubling.

Science users, of course, want a commercial-quality DBMS, i.e., one that is reliable, scalable, and comes with good documentation and support. They also want something that is open source. There is no hope that such a software system can be built in a research lab or university. Such institutions are good at prototypes, but not production software. Hence, the obvious solution is a nonprofit foundation, along the lines of Apache or Mozilla, whose charter would be to build such a DBMS. It could not be financed by venture capital, because of market size issues. As such, support must come from governments and foundations.

It is high time that the United States got behind such an initiative.

 

Reader’s comment:

While I agree that RDBMS is not an optimal technology for scientific applications and that an open source initiative may lead to some good innovation, I’d be cautious in separating the data model from the query and management language.

There are proprietary tools, such as kx.com, that have done so successfully.

The speed and capacity of such tools is phenomenal (as are the licensing fees one must pay).

—Leonidas Irakliotis

from mark Guzdial’s “the importance of Programming in introductory computing courses”

In computer science, the way that we investigate computation is with programming. We don’t want to teach computing as a pile of “ accumulated knowledge.” We know that that doesn’t lead to learning. We need to teach computation with exploration and investigation, which implies programming.

The best research study that I know of that addresses this question is Chris Hundhausen’s study where he used algorithmic visualization in CS1. He had two groups of students. One group was to create a visualization of an algorithm using art supplies. The students were learning theory and describing the process without programming. The second group had to use a visualization system, ALVIS. The students were learning theory and encoding their understanding in order to create a presentation. As Chris says in his paper, “In fact, our findings suggest that ALVIS actually had a key advantage over art supplies: namely, it focused discussions more intently on algorithm details, leading to the collaborative identification and repair of semantic errors.” If you have no computer system, it’s all too easy to say “And magic happens here.” It’s too easy to rely on intuitive understanding, on what we think ought to happen. Having to encode a solution in something that a computer can execute forces an exactness such that errors can be identified.

The idea isn’t that programming creates barriers or makes it harder. Rather, using the computer makes it easier to learn it right. Without a computer, it’s easier to learn it wrong, where you just learn computing as a set of accumulated knowledge (as described in the AAAS report) or with semantic errors (as with art supply algorithm visualization). If you don’t use programming in CS1, you avoid tedious detail at the possible (even likely) loss of real learning.

 

© 2009 aCm 0001-0782/09/0500 $5.00