JISC

The project will develop Search interfaces that insulate the user from the complexities of querying digital repositories using interoperable standards. The open source JAFER software will be developed to provide a simple mechanism of integrating Search and Discover functionality into applications residing in the JISC e-learning framework User Agent layer, namely Learning Design, VLEs, Portals and Resource List Management software. 

Aims and Objectives

The overall goal of the project is to provide robust open source software that insulates developers of academic user agents from the complexities of distributed searching, resource metadata aggregation and reading list creation. We aim to make it as easy as possible to integrate fully functional and interoperable Search and Discover services into Learning Design, VLE and Portal systems.

The specific objectives are to:

• Create Search, Aggregate and Resource List interfaces that can be easily embedded into User Agent layer software (Portals, VLE, Learning Design Systems).
• Provide best practice guides to enable other institutions to customise the interfaces to their needs
• Investigate UDDI schema for Target Repositories

Project Methodology

The project involves developing Java Server Pages that call the JAFER Java classes to make interoperable search queries across multiple distributed repositories.

We will collaborate with Edinburgh University to create Searches across different types of databases e.g. ODBC, SRU, SRW, Z39.50. We will also make queries to different types of metadata e.g. Dublin Core and UK LOM Core.

We will develop software iteratively throughout project and the code will be available on Sourceforge.

Implications

Here we briefly outline how the deliverables will benefit future demonstrator projects in the search and discover context.

• Partnering with FE, HE and commercial organisations to grow a library of customised search GUIs pertaining to the full range of academic disciplines.
• Partner with developers of User Agent applications to integrate the GUIs developed during this project: (a) VLEs e.g. COSE, Blackboard and WebCT; (b) Portals e.g. uPortal; (c) Reading List Management software e.g. Sentient’s Discover product; (d) Learning Design Systems e.g. CopperCore and LAMS; (e) other low-tech integrations such as web sites and office applications like word processing software.
• Working to implement interoperable standards with target repositories such as OCLC, MERLOT, Fedora, DSpace, HLSI, and OAI repositories along with traditional publishers and aggregators.
• Work with SRW and SQI communities to enable peer-to-peer search capabilities
• Integrate the JAFER software to enable OpenURL Resolvers to perform two-way messaging between the Resolver and Target repositories.
• Integrate JAFER messaging with Authentication (e.g. Kerberos) and Authorisation (e.g. Athens and Shibboleth) specifications.

Deliverables

1. A set of fully functional JSP GUIs that enable interoperable distributed searching using Z39.50 and SRW. These GUIs will be customised to the search requirements of the English, music, anthropology and medicine academic disciplines
2. Documentation that describes how developers in other FE and HE institutions can integrate the GUIs produced during this project into their User Agent applications. This will include machine-readable documentation such as WSDL descriptions of appropriate web services and WSDD deployment descriptions for enabling installation of the toolkit web service components. It will also include Ant scripts for building and deploying the software.
3. Documentation that describes how developers in other FE and HE institutions can develop their own customised Search, Aggregate and Resource List GUIs. The documentation will include: (a) technical guidelines on creating GUIs in a Service Oriented Architecture; (b) usability design heuristics for creating search, aggregate and reading list GUIs.
4. Prototype and documentation of a UDDI registry that allows repository owners to enter the details of their database in an interoperable manner.
5. JavaBean that implements the SRW specification.
6. Prototype JavaBean/ documentation looking at the implementation of the SQI specification.
7. JSP GUI that aggregates the metadata returned from the target repositories. This aggregator will allow the user to search through the returned records and select items to feature in the reading list GUI. We will investigate free-text searching and sorting to allow the user to find the required metadata items.
8. JSP GUI that stores the selected metadata items as a resource list. The user will be able to populate the reading list with annotations and add metadata that describes the list for sharing. The list will be compliant with the IMS RLI specification.
9. Functionality that allows the compiled resource list to be downloaded in a number of formats (including one compliant with the IMS RLI specification).
10. Class, User and Sequence UML documentation of all aspects of the project
11. Update version of the existing JAFER JavaDoc documentation.
12. Website for dissemination of documentation, source code and fully functional search, aggregator and reading list GUIs.

Ultimately the project addresses the need of educational institutions to move away from monolithic library systems to a more distributed way of provisioning content to learners. The goal of this project is to create software that can embedded seamlessly into learning environments.

This project is about pulling together the library and VLE worlds.

Project Partners
University of Edinburgh