E-Matrix – NCSU Library E-resources Management System
E-Matrix – NCSU Library E-resources Management System
presented by Andrew Pace and Stephen Meyer
Andrew started by giving an overview of the ERM landscape. DLF wrote a paper on ERMI that describes the problem, outlines existing solutions and efforts, and several appendices. Lots of vendors are working on ERM products and many libraries are developing ERM. There is lots of good work going on at the Web Hub.
After discussion NCSU discovered that ERM really effects everyone. As a result you need to involve all the stakeholders from end-users to serials catalogers. ERM is an attempt to address the shortcomings of existing systems to address the requirements of managing with licensed electronic resources. As a result you end up with a disintegrated system which is a huge inefficient mess.
Goals of the project
- Public interface was a secondary concern.
- Leveraging existing data – all of it!
- Radically Change Technical Services Workflow
- Avoid solutions looking for a problem
- Embrace the serial work
De-emphasize the library catalog. Stop making patrons better library users and make better library tools for patrons to use. Additionally we wanted to put our development money where we were putting our collection money. We spend a large number of $$ on serials/e-resources. Shouldn’t we spent proportional dollars on making these resources available to our users.
Bring together data from disparate systems: statistical data, acquisitions and licensing data, collection management and evaluative data, bibliographic data, local subjects, etc. Must decided where the authoritive data store for information is and make sure that staff conform to this scheme. Lots of descriptive metadata about the databases which is very helpful to end users and reference librarians.
- Container – what kind of resources is it – resource types
- Content – what is inside of it or indexed by – content types
- Aboutness – what is the resource about – web subjects
The end result is that we can create richer reports about our collection and we can faciliate user-centered resource discovery. Part of this are relevancy ranking for databases within a given subject, the ability to add subject-based descriptions, and stable urls for databases (urls are managed in one place and e-matrix acts as a resolver). Dynamically generated database lists that are customized. System also track when resource is down.
Journals are the other component this. The goal is to bring together all the manifestations of a given work. Leveraging existing data in the catalog records, and SFX knowledge base records. Simultaneously display the holdings information for all format types. Journal discovery which reduces the number of clicks. Can assign subjects for core journals in that area.
Stephen did a demo showing a live mock-up which really looked great. The interface was all web-based for staff to use. Staff can view the data in a number of ways. He also demoed a portion of the library’s site which is driven out of the same systems.
If was a terrific session. I learned a lot in the session and got many ideas further ideas about how managing these resources from an administrative standpoint is connected back to how we display them and users access them on the library web site. Andrew didn’t talk much about how this connected to NCSU’s MyLibrary system in the presentation but I was able to talk to him a bit more about that piece of it afterword. Incredibly forward and innovative thinking going on at NCSU.
There are several reasons why this presentation was of interest to me. First, I’d been previously dealing with many of these issues at SUNY Cortland because I was responsible for this area. Second, we are starting the process of examining the effectiveness of the UH Libraries website with particular attention to how e-resource are dealt with and display on the page.
At Cortland, we didn’t catalog any of the e-resource we had at SUNY Cortland and put them in a homegrown system instead. While our system was much simplier than the one NCSU built-in I could see there were many similarities in the types of metadata that we wanted to collect. I talked to Andrew a little about what we had done at Cortland and offered to provide some further information to him if he was interested. Particularly on collecting resource usage data using EZProxy. I have an article on this topic coming out in Library Hi Tech very soon.
If you are at LITA and you missed this session and really want to see it. The session repeats again on Sunday morning.