From NSDLWiki

Jump to: navigation, search



Interoperability of NSDL Collaborative Tools through Aggregations in the NDR


Design Overview


Currently any resource within the NDR is a STEM (science, technology, engineering, and mathematics) resource in the NSDL Collection. This design proposal would lift that assumption and use aggregation membership and projection to create other groupings of resources.

STEM resources selected for the NSDL would continue to be identified by membership in the NSDL Collection aggregation. The purpose of this design is to allow aggregations of resources that are not in the NSDL Collection. Examples of other aggregations might be an NSDL Community aggregation, blog posts aggregation, or OnRamp content aggregation. Resources can be members of any number of aggregations.

Collaborative tools may establish or utilize these aggregations to organize relationships for their own purposes. An important motivation for this would be to search for resources based on membership in an aggregation. Users of a collaborative tool can select items to be submitted as an NSDL STEM resource. Users also can select items to be submitted to multiple aggregations depending on the tool.


Currently, the NDR produces two OAI feeds:

  • Metadata Centric OAI feed for harvest by the public
  • Metadata Centric OAI feed for generating the search index

Both of these are based on contributed metadata for NSDL STEM resources. We will create two new Resource Centric OAI feeds. One is for public use, and will describe only STEM resources. The other will be used to generate a search index that will describe all resources in the repository. This allows all resources to be discovered through the search service.

It is expected that this design will provide the following OAI feeds:

  • Metadata Centric OAI feed for harvest by the public (maintained for backward compatibility)
  • Resource Centric OAI feed for harvest by the public (NSDL STEM resources and OAI distributed metadata)
  • Resource Centric OAI feed of all resources in the NDR for generating the search index


The search index will be generated based on a Resource Centric OAI feed. In addition to resource related search terms, the search index will retain information about the membership of resources in aggregations and the path of all parent aggregations.

The following are searches that are expected:

  • A query can search a selected aggregation. The search will include resources that are members of the selected aggregation, as well as resources that are members of aggregations that are members of the selected aggregation. In other words, searching across a high-level aggregation will also search across any included aggregations. For example, a search across the ExpertVoices aggregation will also search the separate blog aggregations included within ExpertVoices.
  • A query can also explicitly select multiple aggregations. For example, a collaborative tool may use this query to discover resources maintained in other tools in the collaborative tool set through a single search source.
  • A query can search all aggregations, which would include both resources in the NSDL Collection and those that are not.

If no aggregations are specified, the search will be limited to the NSDL Collection of Collections. This maintains backward compatibility, and should provide minimal surprise to the average user. A separate option will be provided to allow searching the entire NDR (something akin to "aggregation=ALL").

Items maintained by the collaborative tools, but not included in the NDR, will not be discoverable through the NSDL Search Service. These items may be discoverable through each tool's individual search facility.

Design Requirements

Collaborative Tools

Each of the collaborative tools will need to have features that interact with the NDR . Issues related to the NDR interaction for each tool are described in the following sections.

Expert Voices

ExpertVoices is an NSDL blogosphere powered by WordPress MU] and custom NSDL ExpertVoices plugins. The blogosphere has multiple blogs with multiple posts related to STEM education topics. Currently selected blogs, posts, and referenced resources can be published to the NSDL Data Repostitory (NDR) by trusted users via the NSDL plugins, making them discoverable in the NSDL Search service and NSDL OAI. The blog post content is managed within ExpertVoices and not stored in the NDR.

We would like to expand the NDR resource to include to blogs and posts which are not in the NSDL Collection. These resources would be members of the blog aggregation and discoverable in a broader search service that other collaboration tools could take advantage of.


NSDL Wiki is a wiki environment powered by MediaWiki] and custom NSDL Wiki extensions. It is a collaboration space for individuals and/or groups to create articles about STEM education topics and document NSDL community efforts. Like ExpertVoices, trusted users can publish select articles and referenced resources to the NSDL Data Repostitory (NDR) which makes them discoverable through the NSDL Search service and NSDL OAI service. The wiki article content is still managed within the NSDL Wiki and not stored in the NDR.

As mentioned above about ExpertVoices blogs and posts, it would be useful to have the wiki articles not in the NSDL Collection discoverable across various collaboration tools for supplemental use, re-use, and expansion.


OnRamp is a content management and distribution system powered by Fez and the NSDL OnFire distribution extension to Fez. OnRamp is approaching its first beta release which will have the basic content management and distribution workflows implemented for destinations that generate RSS feeds. The initial feeds will be used to generate some of the content for webpages within

It is anticipated that release 1.3 will provide the ability to include references to content within OnRamp and results generated by destinations in the distribution system as resources in the NDR. See OnRamp 1.3 Design for more information.


  • Finish MD OAI Exclussion - Make it only pay attention to library area resources.
  • Generate a Resource Centric OAI feed. NOTE: This requires conversion to Fedora 3.0 which is scheduled for release at the end of October 2007.
  • Export of aggregation structure.


  • Search "Describe"
  • Search 3.0 architecture

Use Cases

(work in progress)

User A is writing a wiki article about Global Warning. User B has written several posts about teaching Global Warming to middle school students in ExpertVoices. User B does not have permission to add the posts to the library as resources or does not think they are worthy of being in the library. Currently, the NSDL Search only exposes contributed STEM resources and the Wiki search only exposes other wiki articles. Since User A does not know about or bother to go to ExpertVoices and search, they won't discover the blog posts that have not been added as STEM resources. A broader search that included both STEM and non-STEM resources would allow a more extensible central repository. If User A includes links to User B's posts and think they are indeed "worthy", then User A can add them as referenced resources via the wiki extensions.

Collection A wants to manage their resources and metadata in the NDR, but wants to keep some resources from being exposed until their metadata or content has been reviewed. They could have one aggregation that is a member of the NSDL Collection and one that is not. The one that is not in the NSDL Collection, would not be included in the broader search.

Personal tools