Managing Bibliographic Metadata

Erik Wilde

ETH Zürich

http://dret.net/netdret/docs/wilde-deri05-sharef

Outline

• About me
• Shared References (ShaRef)
  • Project overview
  • Data model
• Questions

About Me (1)

• 50% ETHZ, 30% Lectures, 20% Self-Employment
• 50% ETHZ → Project Leader ShaRef
  • XML-based system and service for bibliographic data
  • Integration with ETHZ CMS and publication database
• 30% → XML Lectures at ETH Zürich and FH Aargau
  • XML and related technologies (XML Schema/XPath/XSLT)
• 20% → Consulting for ISB/eCH
  • Standards and best practices for e-government
  • XML, XML Schema, XML Namespaces, XML modeling

About Me (2)

• History: Document Processing (TUB), Computer Communications (Ph.D.), Web Technologies, Hypermedia
• General interests: Web Technologies and Information Systems
• Activities in different areas:
  • Information Management: ShaRef Project
  • Hypermedia: Text Fragment Identifiers (HT 2005 paper)
  • Content Management: Help/Documentation Tool (HSG)
  • XML for e-Government: Best Practices (eCH-0018)

Abstract

Working with bibliographic metadata is one of the few activities which is shared by researchers of all disciplines. Managing bibliographic metadata is an important activity for conducting research, but only very few researchers use advanced tools or methods for managing their data. In most cases, established tools such as BibTeX or EndNote or even simpler methods such as spreadsheets or text files are being used. The Shared References (ShaRef) project aims at providing a tool and a service for the management of bibliographic metadata. One of the most important goals is to enable the sharing and reuse of metadata. Sharing can be accomplished by using bibliographies which allow access for user groups. Reuse is supported by providing powerful import and export filters, which can be controlled by user options. In addition, ShaRef supports publishing of information by allowing bibliographic metadata to be made available through an HTTP Web Service, which can be used to integrate live metadata into other applications. This talk gives an overview of the ShaRef system and highlights the areas of metadata mapping and metadata reuse.

Shared References (ShaRef)

• Bibliographic data is about knowledge management
  • Bibliographic data as metadata about relevant resources
  • As close as most researchers ever get to formal knowledge management
• Additional data makes the information more useful
  • Personal annotations describing the publication
  • Associations between resources (creates a network of resources)
• Bibliographic data should be shared
  • Much information is managed redundantly
  • Sharing enables research groups to collaborate
  • Personal bibliographies are still possible

ShaRef Usage

• Available through two user interfaces
  • Users are authenticated for access control
  • Java client for online and offline work
  • Web client for browser access
• Java client is an Eclipse RCP application
  • Requires Java 1.5 as runtime environment
  • Connects to the server through Java RMI
  • Local database management using a pure Java DBMS
• Web client is a JSF application
  • Can be used with any standards-compliant browser
  • Connects to the server through Java RMI

ShaRef Data Management

• Data is stored in bibliographies
  • The unit for access control
  • Comparable to BibTeX or EndNote files
• Bibliographies contain entries
  • References and shadows as the main constructs
  • Definitions for keywords and associations
• Import and export facilities are available
• Data can be published through channels
  • Views of bibliographies
  • Published using an export format
  • Access through a URI using HTTP GET

ShaRef Data Model

Supported Input and Output Formats

• Supported Input Formats
  • BibTeX
  • EndNote
• Supported Output Formats
  • BibTeX
  • EndNote
  • HTML (for Web publishing)
  • Silva XML (for CMS integration)
• Implementation of additional formats
  • XSLT-based transformations can be introduced easily
  • XSLT 2.0 supports parsing of non-XML formats

OpenURL - The Link to the Library

• OpenURLs are context-sensitive links for library resources
  • HTTP URL with metadata as query string
  • OpenURL resolver returns the holdings for the resource
• A quick path from personal data to the library catalog
  • Requires the configuration of the OpenURL resolver
  • Depends of the configuration of the OpenURL resolver
• Enables integration of personal and library catalogs
  • Currently only supported for HTML export and publishing
  • Will be integrated into the Web client's interface

HTTP Web Service

• Machine-readable data available via HTTP GET
• Data is cached and refreshed periodically
  • XSLT (2.0) transformations are not very fast
  • Most data is not time-critical
  • If necessary, a refresh can be requested
• Configuration through the Web user interface
  • Pro: Well-known interface for ShaRef users
  • Con: Configuration change requires UI access
• If there is demand, a SOAP interface will be provided
  1. SOAP-based query of bibliographies with subsequent transformation
  2. SOAP-based creation and management of channels
  3. SOAP-based management of bibliographies and entries

What about RDF?

• ShaRef is about metadata - why not use RDF?
• The short answer: Nobody asked for it!
• The long answer: XML is better supported than RDF
  • Schema languages exist for both languages
  • There is no equivalent of XSLT for RDF
  • There is no equivalent of XPath for RDF
  • Java has better XML support than RDF support
  • Most existing formats are text or XML → easy to map
• New formats can be added easily
  • If somebody asks for RDF, we will support it

What about MARC? Or MODS?

• What are MARC and MODS?
  • MARC is the most popular format for library catalogs
  • MODS is a subset of MARC (not all MARC fields are supported)
  • The MODS XML Schema uses language-based names
• ShaRef is about metadata - why not use MARC/MODS?
• The short answer: nobody asked for it!
• The long answer: MARC/MODS are overly complex
  • Deeply nested fields rather than mostly flat structure
  • Not really embraced by anyone outside the library community
  • Only few tools support import/export of MARC/MODS
• New formats can be added easily
  • If somebody asks for MARC/MODS, we will support it

Conclusions

• A new tool for bibliography management
  • Integration into the ETHZ IT landscape in 2006
  • Expected feedback from users and integrators
• Interesting case study for XML-centric software
  • Existing XML technologies and tools are nice
  • Missing pieces in the system development process
• Mash-ups as the new breed of applications
  • Data should be made available in a simple format
  • If it is useful, it will be reused

Thank You! Questions?

• ShaRef Project
• My Web page: http://dret.net/netdret/
• My Email: net.dret@dret.net