Two new visualizations for ELAG

I'm doing a presentation on visualizing library data at ELAG this week. I'm showing off some demonstrations that I built for code4lib 2011 and some new demos that I've built for ELAG. The new demonstrations for ELAG take advantage of two OCLC web services.

The first demonstration leverages the Classify web service. You may or may not be away that the full response from the Classify web service includes a listing of all the editions for a given work, the number of libraries holding that edition and the format of that edition. What that enables me to do is see all how many libraries hold a particular Work in a particular format. So I'm building a pie chart that shows all the holdings for each format.

The second demonstration leaverages the WorldCat Registry web service. Since the Registry contains data about libraries and their branches, it is possible to build an organization chart that visually shows the library and its branches. One issue with this particular code was the fact that the main library record is often represents both the main institution and the main branch of the library that needs to be representat. However, this isn't always the case and it is difficult to determine when this is true and when it is not. So right now the code simply considers the "branches" libraries that have a seperate record in the Registry which is linked to from the main record.

I wanted to develop another visualization but didn't have the time. This was the idea of creating a chart that represented how many peer institutions have a given Work. This could be done by using the WorldCat Search API and a string of all the OCLC symbols that are peer institutions. Sending a holdings request to WorldCat Search API with all the OCLC Symbols of the peer institutions would tell you how many of them have the Work. You could then build the chart based on this information, the number of peer institutions and the Google Visualization API. Hopefully, I'll be able to get this code written and up early next week.