Melissa Wadman and I were on the road again last month, but this time around we visited Atlanta to talk to attendees at the American Alliance of Museums Annual Meeting about the strategies that we have taken to improve the discoverability of Smithsonian-authored learning resources.
That's right, we were talking about metadata. You know, that “structured information that describes, explains, locates, or otherwise makes it easier to retrieve, use, or manage an information resource. Metadata is often called data about data or information about information (National Information Standards Organization: Understanding Metadata). Specifically we were sharing the results of a two year project to develop educationally-relevant metadata to describe the more than 2,000 digitally-accessible learning resources produced by more than 30 agencies (museums, libraries, research centers, etc.) here at the Smithsonian. Those resources, which are currently indexed and searchable at smithsonianeducation.org, originally were described with only basic descriptive metadata (title, description, grade level). Working with an external vendor (who in turn worked with a large team of content specific classroom teachers), we greatly expanded the scope of that metadata following the specification developed by the Learning Resources Metadata Initiative (LRMI). LRMI is part of an even greater project called Schema.org (created by Google, Bing, Yahoo, and Yandex) which strives “to create, maintain, and promote schemas for structured data on the Internet." Schema.org describes all kinds of content you many find via the Internet, but does it in a structured and standardized way so that the search engines can understand the nature of the content they find on webpages.
Schema.org LRMI Fields:
General Terms (schema.org)
The ultimate goal of schema.org, and of the Smithsonian LRMI project is to help end users (teachers, in our situation) find the things they are looking for, and eventually, to provide tools that make search, via the participating search engines, more customizable (see the potato salad recipe example below in which Google has provided additional search filtering tools for webpages that have metadata describing their recipes following the schema.org standard, and imagine how something similar might help teachers filter digital learning resources.)
And the Smithsonian is not the only museum involved in this kind of metadata work. The North Carolina Museum of Art, on its fantastic arts integration program for K–12 teachers, ArtNC.org, has developed LRMI metadata to describe its more than 60 model lesson plans. Their goal too was to make their lessons as findable and usable as possible. They accomplished this through first adjusting the ways that they described their materials to better fit the LRMI standards, and then integrated a dropdown-list-based tagger into their website's content management system. This allows them to quickly assign relevant metadata to new resources as they are published.
Another outcome of having richer descriptions for its content is the Smithsonian's ability to distribute that metadata via the federal Learning Registry, a joint effort of the U.S. Department of Education and the U.S. Department of Defense. The Learning Registry is simply (and this is a big oversimplification) a big pool of metadata that describes learning content from a variety of sources (like OER Commons, Library of Congress, the National Archives, PBS). That pool is open and available to any website developer that wants to aggregate and recommend resources to their users. For example, the departments of education in states like California (Digital Chalkboard) and Illinois (Illinois Shared Learning Environment's OER site) provide portals for teachers in their states to access relevant content. These portals contain search engines of suggested resources. By sharing our metadata into the Learning Registry, they are available to these website developers, and we can ensure that our resources are discoverable to the teachers who use them. Check out our presentation slides for some examples of Smithsonian resources surfacing in these portals.
Along with Flight Design, we also put together this infographic to help other museums and similar organizations who might be considering a metadata project of their own (download a higher resolution PDF version to hang on your fridge).
At the conclusion of our presentation in Atlanta, we provided a few recommendations for museums specifically:
Museum educators can
- Create LRMI metadata for their existing resources.
- Work with digital staff to embed LRMI metadata on resource HTML pages.
- Develop LRMI for emerging resources as they are developed.
Museum technologists can
- Assist museum educators in utilizing tagger tools.
- Work with museum education staff to embed LRMI metadata on resource HTML pages.
- Publish LRMI metadata to the Learning Registry.
- Play an active role in the LRMI and the Learning Registry communities to ensure that the needs of their institutions are well represented as these initiatives continue to develop and mature.
So, how will all of this make its way into the Learning Lab? Well, as the Lab enables more and more educators to personalize Smithsonian-authored learning resources and to create their own digital learning resources, we might have ourselves a problem. At launch, users will find themselves searching through more than 1.3 million digital resources from the Smithsonian. As we and our users add to this, the problems of searching through such vast resources will only multiply. To attempt to stem this problem, we will be building a tool that allows users to tag the resources they have modified or created using the LRMI framework. The wireframe below (more on wireframes coming soon) demonstrates one way in which this might work. As our users go to publish their creations (so that others may find and use them) they will be prompted to supply some metadata (we hope lots, but we'll settle for a little). Emphasis will be placed on encouraging them to supply at least the target age range and area of study, as these are the two most important metadata fields for teacher analysis of potential learning resource (according to our own research and several recent surveys by the folks that created LRMI).
We still have a lot of unanswered questions about this tagging on the Learning Lab: will our users take the time to add this to their resources? Should the metadata describing resources published on the Lab be shared into the Learning Registry? And it goes on.
I'm pretty excited to find the answers!
Image: Two Men Uncovering Fossil
Image of two men uncovering a fossil specimen called Sp. 20-27. The caption on the back of the image reads, “Photo-0-sp. 20-27. Lower jaws are seen to the right of the picture with teeth well preserved. Corse sand Lower Oligocene beds. Wyo." Scientific field research headed by Charles W. Gilmore, curator of vertebrate paleontology for the U.S. National Museum (USNM), now known as the National Museum of Natural History, was conducted in 1931 and 1932. The expedition covered the Miocene and Oligocene formations of southwestern Montana and the Bighorn Basin in Wyoming. Expedition members sent back fossils to the USNM for research and display. Smithsonian Institution Archives.