IKE in 3D

I’ve been playing with inductive clustering of digital reference questions. To this point I’ve been visualizing this in 2D, but I have been keeping track of three dimensions.

I’ve discovered a great little development application distributes as part of it’s developer package called Quartz Composer. It uses a visual programming language to directly tap into Apple’s graphics architecture. It’s inputs aren’t too rich, but I managed to get MySQL data dumped into it using RSS. While it seems a bit of a kludge, it works pretty well.

Here’s an animated GIF showing the digital transactions clustering in 3D:

3D Ike

Here is a quick MPEG movie showing how you can fly around the cluster at any point. What you’ll see is a screen capture of me moving around using the keyboard and mouse:

3D Cluster Move

Here’s just a pretty rotating movie of a small subset of the data:

3dplay.mov

Reference Authoring

“Reference Authoring” School of Information’s iForum, University of Texas, Austin, TX
Wednesday, February 8th, 2006

Abstract: Digital reference is not just reference interviews online. The main difference is the production of a “reference artifact.” In face-to-face reference, work must be done to retain the transaction (it must be recorded, or written down…action must occur), whereas in digital reference the opposite is true (the e-mail must be deleted, the database purged). This may seem like a small difference at first, but it is critical. With recorded transactions knowledge bases can be created, pathfinders authored, training can occur with real data, etc. This use of digital reference output is called reference authoring. This presentation will explore reference authoring, the use of induction and complexity research to manage knowledge created through reference authoring and the increasing intersection between reference and information retrieval.
Slides: https://davidlankes.org/rdlankes/Presentations/2006/RefA.pdf

“Reference Authoring” School of Information’s iForum, University of Texas, Austin, TX

Slides now available:

Digital reference is not just reference interviews online. The main difference is the production of a â??reference artifact.â?? In face-to-face reference, work must be done to retain the transaction (it must be recorded, or written down…action must occur), whereas in digital reference the opposite is true (the e-mail must be deleted, the database purged). This may seem like a small difference at first, but it is critical. With recorded transactions knowledge bases can be created, pathfinders authored, training can occur with real data, etc. This use of digital reference output is called reference authoring. This presentation will explore reference authoring, the use of induction and complexity research to manage knowledge created through reference authoring and the increasing intersection between reference and information retrieval.

IKE Redux

A while back I posted on a project called IKE (Inductive Knowledge Engine) that uses complexity theory to inductively organize digital reference knowledge bases. I’ve been back at work with it, doing a substantial rewrite of the underlying system. I’ve also spruced up the look of the IKE output (including a cloud view, but that’s for later). I thought I would share the most recent output.

NewIKE.gif

In this figure 200 question/answer pairs from the AskVRD service are shown clustering. This is a very deterministic test run with 1 simple rule applied to all items…get closer to items that share your subject. This deterministic test is used to make sure all the software is working as expected. The next step is to include multiple rules, then dynamic rules (like how popular is this subject), and then finally diverse agent types. Still, I though the picture was pretty enough to share.

In case you want to know what the actual rule involved looks like (in PHP):

foreach ($prime as $p){
foreach ($target as $t){
if ($p[0]==$t[0]){
if (($p[1]==$t[1])AND($p[0]==”category”)){
$Move++;
}
if ($p[1]!=$t[1]){
//$Move–;
}
}
}
}

…and Whatever

The following was a letter I wrote for a party celebrating Mike Eisenberg’s achievements as a dean at the University of Washington. If you know Mike, read on. If you don’t, this probably won’t make much sense.

AEGang

A Salute to Michael Eisenberg
Dean Emeritus

By R. David LankesI have known Mike as a boss, a mentor, a colleague, a business partner and a friend. I owe a great deal of my success to Mike. He always took the time to help me, and guide me. It is telling, however, to realize that much of this guidance came at street corners, U.S. Airways Clubs, and in cab rides. Scott Walters talked about Mikeâ??s â??drive by advisingâ?? style. It made me remember that Mike and I actually agreed on my dissertation topic at a Skychiefâ??s baseball game.

But Iâ??m convinced that the key to Mikeâ??s success is not what he says, or even where he says it. His success is in what he doesnâ??t say…specifically what he hides behind the phrase â??and whatever.â??

Many are the times Iâ??ve been in a meeting with Mike where he begins speaking, begins a sentence, and then ends with â??and whateverâ?? as if we all know what he skipped. â??Weâ??re going to create a new Internet service called AskERIC. It will answer teachers questions…and whatever.â?? â??So your going to do a dissertation on complex systems where you study and whatever.â?? â??Ok, so youâ??re going to give us a million dollars and weâ??ll…whatever.â??

For the longest time I thought Mike was thinking faster than his mouth would work. That locked behind â??and whateverâ?? was a complex series of plans, details and thoughts. That he had simply played out the entire conversation in his mind, much as a chess master can look ten moves into the future, and that he couldnâ??t be bothered to translate those plans into words. Then, I realized his true brilliance…he had no idea what came after the first part of the sentence.

What a scam! He would state some titillating and dramatic idea and then throw in a â??and whateverâ?? allowing the listener to fill in the details. It is like a bizarre verbal Mad Libs game where he hands you the story with massive blanks and has you fill in the rest. The audience choosing what is best for them, finds it a perfect match to their needs, and then attributes the results to Mike. All he has to do is prompt you for an answer and take credit…brilliant.

Think of the implications. He could receive the Nobel Prize in medicine with his daring work on stem cells, summed up in his 2007 paper â??Curing Cancer with Stem Cells by…and Whatever.â?? He could win the Newbery Medal for Children’s literature for his best selling book of one page entitle â??Little Red Riding Hood went into the woods and Whatever.â??

I attribute this devious methodology to Mikeâ??s early work in relevance. I have no doubt that Mike quickly discovered that the less actual content a document contains, the wider the potential relevance of the document. By simply sprinkling a document with grand pronouncements and not being weighed down by details, or reality an item can be seen as relevant to a mass of unsuspecting user population…letâ??s call it the Eisenberg Coefficient of relevance…the fewer the facts, the greater the appeal. No one could fault Mike for this discovery, we just wish he hadnâ??t shared it with the Republicans.

This of course brings me to some of Mikeâ??s other, lesser known contributions to the information science literature. There is the ongoing experiment into how numeric qualifiers can be used to achieve economic gain…Big Six, Little Twelve, Super 3…word on the street is that in following Appleâ??s iPod Nano success, the Microscopic 24 is on the way…and he may well copyright the Big Red One.

And who could forget Eisenbergâ??s Model of Funding Scaffolding Saturation. Mike empirically proved that you can get more money out of a funding source by constantly adding costs until the funder simply gives up:

Mike: We can do that project for you for $300,000
Funder: That sounds reasonable.
Mike: Then thereâ??s overhead.
Funder: Of course.
Mike: And travel.
Funder: Thatâ??s not part of the 300k?
Mike: No
Funder: OK
Mike: That will be for 6 months.
Funder: I thought a Year…we really donâ??t have 600k for a year
Mike: How about 400k, we could probably get by on 400k for the year
Funder: I guess so.
Mike: Of course that will add to the overhead and travel…

The list could go on, but I will end on my favorite of Mikeâ??s hidden theories, The Saint Peterâ??s Test of Success that followed up on Katzerâ??s Law on Change…â??change is like heaven, everyone agrees it is a good idea, but no one wants to go first.â?? The Saint Peterâ??s Test from Mike is surprising considering both Mikeâ??s quantitative background and the fact that he is, well, Jewish. In this test one comes before Saint Peter standing before the Gates of Heaven. Saint Peter asks you â??Well, did you leave it better than you found it?â?? It is a seemingly simple test, but it does tend to simplify the questions around what is success. Perhaps it is also fitting to talk about this test at this point in Mikeâ??s career. For indeed, with Mikeâ??s work in schools, his leadership in information, his impact at Syracuse, and his amazing legacy at the University of Washington, Mike has already passed the Saint Peterâ??s Test with a lot of room to spare. I look forward to seeing what Mike does next.

PHP Podcast Online

I have added a Podcast session that does a simple introduction to PHP. For those interested (or any other episodes in my podcast) the RSS feed is at:

https://davidlankes.org/rdlankes/pod/pod.xml

New Web is Ready

The new home page is done. Well, mostly. I am still adding older (pre-2003) presentation archives over the next weeks, but most of the material is in place.
Some have asked me why I redid my home page. First, I’ve been doing a lot of simplifying of my information life. My old web had grown a bit unwieldy over two years. CMS systems like TikiWiki are great, but maybe a bit too much for what I needed. There was also a constantly security concern with keeping patches up to date and the like. My new site is high on function, but slim on features: a blog, a search engine…pretty simple. It is also easier to update with new publications and presentations. All I have to do now is drop a file in a directory, and the index will be updated. No menus to edit.

For those who care, the two pieces of technology at play here are Nutch, and opensource Java based search engine, and WordPress for the blog.

Always looking for feedback.

New Podcast

I’ve uploaded a new Podcast. This one is is the audio portion of a LAMP Overview I did for my digital library class. I will be following this up with introductions to Linux, Apache, MySQL and PHP.