Explore Open Archives-Metalist | Evaluation | Tools and resources The Open Citation Project - Reference Linking and Citation Analysis for Open Archives

 

Research and development

On this page: Architecture for reference linking | Services: finding papers in Open Archives | Software solutions | User studies: mining Web logs and user surveys | OAI and metadata | Early OpCit

This page provides access to selected documentation on early-stage research by the project.

The main formal dissemination route for the results of our research will be through published papers and presentations, or in the form of demonstrators designed for evaluation.

Architecture for reference linking

• An Architecture for Reference Linking (Cornell OpCit site), Donna Bergmark (this version May 2000)
  Why a new architecture? "True to Cornell Digital Library Research Group tradition we prefer an object-oriented approach. ...in our object-orientation, we want to walk up to a paper and ask it ‘what are your references’ or ‘what is your title’ or ‘what is your DOI’? This led directly to inventing the concept of constructing a surrogate for each paper, specifically for each Item in an archive that is being analyzed. The surrogate would know how to answer these questions. And which questions in particular it knows how to answer is called the API". See also Reference linking working notes (Cornell OpCit site)

Services: finding papers in Open Archives

• Citation-ranked search engine for eprint archives: Citebase Tim Brody (2001- )
  Building on OpCit's bibliometric analysis work citation database, Citebase will continue to be developed by Tim Brody beyond OpCit. Try the prototype search engine. Indexes arXiv eprint archives.
• OAI Aggregator: Celestial Tim Brody (2002- )
  Celestial is an experimental service that aims to offer a comprehensive and accurate cache of up-to-date records from known OAI repositories. If OAI service providers harvest from this service, the load on data providers will be reduced. An effective aggregator will improve the interoperability, scalability and reliability of OAI services.
• Visualisation of simple citation and co-citation data. Two mapping tools have been developed:
  • OpCit e-Services Simon Kampa (2002)
  • XGQuery (Cornell OpCit site) Xiang Guo and Donna Bergmark (2002)
  Currently these services can only be shown graphically. The processing power required to run these applications is too high to present as Web applications.

Software solutions

• ParaTools, software tools for reference parsing. ParaTools includes two modules - the former provides functionality based on the Citebase citation parser, and the latter provides an accurate means of reference extraction from documents - written by Tim Brody, which are updated versions of the original reference processing software produced for OpCit by Zhuoan Jiao.

User studies: mining Web logs and user surveys

• Bibliometric analysis: mining the social life of an eprint archive Tim Brody, Ian Hickman (July-September 2000)
  Authors, publications, citations and impact
  Usage patterns: Web log analysis, site hits and downloads
  The Los Alamos physics archives have been accepting submissions since 1991. Why have eprints not replaced the traditional publication process? What is the relationship between the publication process and the submissions to eprint archives? Do author's receive better exposure by submitting to an archive? How does peer-review affect the eprint archive?
  See also Early Opcit: Preliminary user/citation analysis of arXiv, Les Carr (July-November 1999)
• A survey of users and non-users of eprint archives Cathy Hunt (2001)
  Who is and is not using eprint archives, how are they using them, why they do not use them, and what features should be added to make the archives more useful. Results and report from the survey.

OAI and metadata

• Full text services within the Open Archives Initiative: an OpCit proposal. Simple metadata mandated in the OAI protocol is not sufficient to build a reference linking service, which requires access to the full texts of the papers. OpCit proposed A Storage Architecture for Full-Text Access to Open Archives. As the OAI has become esentially a protocol, it is unlikely to support such developments directly. Instead, OpCit is working with the arXiv data provider to test software that will allow data providers to extract references automatically from papers as they are submitted and then present these data to service providers.
• Academic Metadata Format (AMF)
  Having harvested citation data from arXiv, OpCit is looking for an output format so that it can make its data available to other OAI data and service providers. In collaboration with Thomas Krichel and Simeon Warner, AMF is one possibility being investigated.

Early OpCit

• Early linking experiments, Les Carr (July/August 1999)
  From recognising an initial set of within-archive citation links and a set of explicit arXiv citations containing an archive reference number, to overlaying these links on the original documents in text format and then in PDF.
  • First link demos on arXiv
  • Report on early link demos and directions
• Copying arXiv? Zhuoan Jiao (last updated May 2000)
  Converting arXiv to PDF: "As a viewable, linkable format, PDF is convenient for presenting a linked archive. The majority of the papers submitted to the arXiv physics archives are in other formats. PDF files are permanently stored and do not require the user to wait for conversion".