archives
Identify theft experience highlights importance of use restrictions
The original appearance of this entry was in Danny Weitzner - Open Internet Policy
The actual harm done by identity theft has some important clues for those concerned about privacy protection in environments where information flows around through multiple channels. In an excellent account of an identify theft victim’s experience (which began well before the Web, viruses, spam, etc.), Tom Zeller of the New York Times (Waking Up to Recurring ID Nightmares - New York Times, 9 Jan 2006,) illustrate the need to focus on controlling the useof information, rather than trying to control collection Zeller quotes a source from an identify verification company:
“They say once the horse is out of the barn, why bother closing the door?” Mr. Waller said, referring to the millions of bits of consumer data already leaked into the black market. “But even if someone has your Social Security number, if you can prevent them from using it, that’s the solution we should be driving towards.”
There certainly is some information that should never be collected in the first place, but this story illustrates that a key aspect of privacy protection in more open information environments will be creative mechanisms for controlling, both legally and technically (following the users own choices and preferences). I’ve been trying to understand how to do this in recent speaking and writing. This is all still a work in progress.
Data mining for (and about) the rest of us
The original appearance of this entry was in Danny Weitzner - Open Internet Policy
The applefritter blog has a very clever entry showing how easy it is to data mine information publicly available on the Web in order to learn all sorts of revealing things about people. Tom Owad shows in his post, Data Mining 101: Finding Subversives with Amazon Wishlists | Applefritter how it possible to piece together relatively simple web technology to create detailed profiles of people who, in this case, have public wishlists on Amazon.com. Wishlist data, combined with other services (Google maps, geocoding tools, etc.), can be combined to give of view of what sort of books people are interested in reading and where they live. (I was at first a bit surprised that wish lists are public by default, but it makes some sense. If you have a wish list you want people who might fulfill your wish, to have access to it.)
Owad hastenes to add that his exercise is not meant as an attack on Amazon, but rather his goal is to underscore the importance of putting limits of how the government can actual use the information that they gather from a variety of sources. I agree with his view and have said so elsewhere, but think that this relatively simply hack proves far more. Owad’s hack (a term I use with only its positive connotations) points out the now widely-dispersed power to infer a lot about individuals based on information that we all leave behind on the Web. As web services, AJAX and Semantic Web technology become more popular on the Web, this trend will only increase geometrically. Some react to this by looking for ways to minimize the trail of data we leave. I believe that to expect people to do that is neither fair nor realistic. We want to encourage user services like wishlists and innovative uses of that data beyond it’s original intent, provided that these uses don’t contravene the data subjects expectations and provided everyone feels confident that there is a legal framework that protects against abuse. As we begin to rethink privacy laws for both the government and private sector, we should think more about how to set generally-accepted rules about how data will be used.
Arpeggio in D, a little three chord ditty
I ran across Ping on improvisation the other day. It seems he learned to play mostly from sheet music. I mostly learned by playing in a music group at church. I learned a few chords in a classroom setting, but mostly, I would sit down in the church group and George or Rudy would lead the group and I would try my best to follow. My crowing achievement was one day when neither of them was available; it was just me and a gal on flute, and we pulled it off. Unlike my friends with real talent that learned to play better than I ever will in their first year, it took me at least five years to achieve that level of competence on folk guitar. I need to hear the song and see the chords before I can play it. My ear training is proceeding very, very slowly; it took me years to learn to tune my own guitar.
I have picked up several guitars over the years, but it was years of wishing before we got a piano for our house this year. It was out of tune enough that I could tell, and I had to leave it that way for a month while it settled. On the day of the tuning appointment, I was tidying up the piano room a bit and I couldn't help but sit down and plunk around a bit. The piano tuner came in and asked if I was the piano player in the house; I said no, not really; my son was taking lessons; I just fake it, using my basic three-chord guitar sense. I was relieved that he didn't sneer at this approach, but rather agreed that they should teach chord progressions and the like to beginning piano players. In Ray, there's a flashback that shows him learning more that way.
Anyway, I found Ping's piece on music just as I was running out of steam for technical work, so I headed down to the piano and worked on a few of the easier pieces of my guitar music. I goofed or got frustrated with one or something... and then I wandered into this 1-5-1 bass arpeggio* thing in D... first just I/IV I/IV... then before long the V chord (A) shows up... and after that got monotonous, a Bm bridge showed up. And then I could hear a melody in my head. I can't play well enough to do both the melody and the bass line at the same time, but going back and forth, I sorta worked it out: 
On the one hand, it's so simple that it's sort of embarrassing to call it an original composition. But it's not every day that my muse visits me this way, and I'm so in the habit of sharing in the Web that I started thinking about all the issues around music markup in the Web.
I'm not talking about mp3 vs ogg; I'm talking about sharing something editable:
There are very few data formats I trust... when I use the computer to capture my knowledge, I pretty much stick to plain text, XML (esp XHTML, or at least HTML that tidy can turn into XHTML for me), RCS/CVS, and RFC822/MIME.
I use JPG, PNG, and PDF if I must, but not for capturing knowledge for exchange, revision, etc.
GarageBand is a blast; I'm really afraid of becoming addicted to it and locking up all my music in there. Version 2 has support for western music notation. Plus, it lets you record tracks separately and mix them. So I gave that a whirl; you can listen to
arpeg-d.mpg,
mistakes and all;
but there doesn't seem to be any way to get the music notation out of GarageBand. The extraction of data created in GarageBand does not appear to be an easy task
-- Dent du Midi FAQ.
This is not the first time I have been in this position; I wrote a few songs in college and transcribed them on my Macintosh SE circa 1988. When I recovered the files a couple years ago, I searched for a more modern format and found ABC music notation that is editable and convertable to postscript sheet music and MIDI; fortunately, the Studio Session format documentation survived and I could write a a python ditty to convert my data.
So tonight I captured Arpeggio in D for sharing:
... and a makefile to tie them together. I haven't really decided on the melody for the bridge, and abc2midi doesn't grok the bass cleff extension so the bass should sound two octaves lower. But there you have it.
* reading up on musical terminology, I see that I'm perhaps misusing the term arpeggio; it's really a broken chord.
see also: advogato item, notes on debian linux and music tools
