Data Unbound

new IT architecture books

Originally uploaded by Raymond Yee.

Today, after receiving my book order of Understanding Enterprise SOA and Semantics in Business Systems: The Savvy Managers Guide from amazon.com, I set out to start re-reading them in earnest. I'm looking to them to help shape my understanding of service-oriented architecture and the infrastructure we need to build for better data services on campus. There's lot of hype around SOA; I'm trying to distill some concrete actions I can recommend for my colleagues and me to take with respect to SOA. Furthermore, how can I explain this topic quickly and effectively to people throughout the organization, specifically for my context, which is higher education.

I often turn to Educause for a specifically higher ed take on information technology. Lately, I've started wondering how other universities are responding to all the talk and much hype around service oriented architectures. A good place to start is the Service-oriented Architecture section on the Educause website. I've printed out the slides from a talk on SOA by Jim Phelps of University of Wisconsin-Madison earlier this year. Has UW-Madison actually successfully implemented a SOA process?

A friend just asked me how the trackback mechanism actually works. I'll refer him to the trackback technical specification
but want to link to an entry in my JazzUpTheWeb blog to see whether a trackback shows up.

later: yes, the reference worked but I learned that it's a pingback, not a trackback, that is being used in wordpress.com.

I'm reading Dave McComb's Semantics in Business Systems: The Savvy Managers Guide, thanks to my colleague Thom King. Skimming the book left me with the impression that I can grok the essential message of the book but I won't be able to get to the real depths unless I carefully re-read the book. To that end, I begin at the end of the book with Chapter 15 ("Getting Started"). Two lines strike me:

At this point you should be experimenting with OWL and RDF, Web Services, and message brokers, if you haven't already done so. You should also been experimenting with some of the more abstract concepts brought up here, such as semantically inspired elicitation and data profiling as a means to find meaning."

and

build a message model as the basis of your integration efforts

As a data architect, though I'm usually focused on the backend systems, I also deal with clients. I'm also part of a larger organization that deals with with many customers beyond my area of work. I will have to refer the folks designing our customer service operation to a recent article by David Pogue, Business 101: Quality Customer Service Breeds Customer Loyalty - New York Times:

What are they, nuts!? They are actually *inviting* people to call them for free technical support? Don’t they have any idea how that idea will kill their revenue stream? Haven’t they learned anything from the computer industry?

Now, this isn’t really a review of the Pioneer AVIC-D2, although I’m generally happy with it. [....]

No, this is actually a review of the company and its policies. The fact that 25 readers independently wrote to recommend Crutchfield tells me that this hyper-service-oriented approach has succeeded; the company has essentially cornered its market and generated a massive audience of rabid and repeat customers.

I have found the phone service of most companies I've dealt with to be uninspiring. IBM has (or had) highly competent support for its Thinkpads. Apple Computer hasn't treated me that well with a close-lipped, pseudo-friendly attitude. Now, Konica-Minolta's support for its color laser printers has impressed me thoroughly. Each of the several times I've called (save for one instance) has gotten me a friendly, knowledgeable, straightforward, and helpful service agent.

From Are You Ready for Enterprise Architecture?, we find wise words on expressing the value of IT architecture to our constituents:

The purpose of establishing IT standards is to reduce complexity and increase efficiencies through consolidation and improved collaboration. You will find that standards enable an IT team to do more with less. You can cut costs significantly by consolidating data centers, standardizing platforms, and pruning redundancies.

However, you should be aware that EA groups are frequently considered cost-overhead and as such are subject to cost-cutting purges. To avoid this problem, you should focus initial EA initiatives on consolidation projects that dramatically cut costs. [emphasis mine] Then you need to communicate your EA successes and the value EA has achieved with other groups in your organization.

On July 18, 2006, I officially took on a new role at UC Berkeley. As part of a reorganization, I became a data architect working for the Data Services Department of Information Services and Technology at the University of California, Berkeley. Only the day before, I was the Technology Architect for the Interactive University Project. In some ways, there hasn't been much change. I sit in the same cubicle (though staff might be physically moved around starting next year), report to the same direct supervisor, work with some of the same people, and grapple with some of the same challenges. In other ways, great changes are afoot. I obviously don't know what all of them are, but that won't stop me from reading the tea leaves to figure them out.

One change for me is to retire the IU Technology Architecture Lodge, a weblog that I have maintained more or less continuously from March 2, 2000. By retire, I don't mean take the materials off the web. No, I'm actively working out a strategy for preserving the content and migrating the materials from my blog to another writing space. The IU Manila server will be eventually turned off, and we have offered to provide statically rendered versions of their sites to owners. If I come up with any better solutions (which I am currently developing on my own time), then I plan to share them with the public.

An obvious question for my readers is what is going to happen to the work of the Interactive University as a whole. I can speak only to the work which which I've been most directly involved, namely the Scholar's Box and the IU/CDL Collaboration. (I suspect that there will be official announcements about the IU as a whole.) For a long time now, my colleagues and I have spoken about putting the Scholar's Box out as an open source project. I'm pleased to say that we are finally making crucial progress on that front. I have now submitted the Scholar's Box for approval by the Office of Technology Licensing. Perhaps within a couple of weeks, I will have the go-ahead with officially annoucing an open source project around the Scholar's Box.

I hope to find some way to continue a collaboration with the California Digital Library. Given the mandate of Data Services ("Provides stewardship of both academic and administrative data, as well as tools for data presentation, visualization, analysis, and collaboration. Helps prepare the campus for the convergence of IST, Library, ETS, and appropriate external providers."), it would make sense of doing more work together. I would like to write more about issues of data architecture and digital libraries.

So what will this weblog be about? I will write about issues related to the work I do as a data architect at UC Berkeley. I anticipate writing about such topic as:

• data architecture–what it is and how one does it
• the relationship between data architecture and other forms of information technology architecture
• research, teaching, and administrative data in higher education
• cyberinfrastructure
• service-oriented enterprise architectures
• what we are doing specifically at IST, UC Berkeley, and the UC system as a whole
• IT organizations
• IT staff development
• scholarly data
• scholarly workflow automation
• digital libraries
• educational technology
• collaboration, presentation, and analysis tools