View
215
Download
0
Tags:
Embed Size (px)
Citation preview
NEEO Workpackage 5
NEEO Project Meeting - 5Geneva, Switzerland
22 June, 2009
Benoit PAUWELS
• Agreement– decision to outsource portal design– financing (12.000€)
• partly from NEEO (unbudgetted) and Nereus
• BlackAndWhiteCompany – Kortrijk, BE– http://www.blackandwhitecompany.be/
EO portal
• Specifications– Input:
• IWG – 2 meetings• Remarks on graphical design of first version
of portal• Results from mid-project URR • Additional functionality (export, download
statistics, MLIA, full-text searching, enriched metadata, usage reports)
– Written specifications: BP and Cécile (ULB)• Available on SurfGroepen: here
EO portal
• Implementation– Several rounds of proposals, remarks, amendments,
based on delivered JPEG sreenshots: IWG
– Close communication between B&WCompany, BP, FV
– Last-minute requests: voted (IWG + funders)
• Result– All screens are available as HTML/CSS/Javascript– New portal available: will show it in a minute
• Will need revisiting– datasets– MLIA– …
EO portal
• Review your institution and authors pictures– Technical guidelines 1.4 – Annex 3
– Institution logo and thumbnail
– Author photo
EO portal
Type Maximum height in pixels
logo 65
thumbnail 33
Ideal height in pixels Ideal width in pixels154 105
• Roadmap
Version 1.4 (2009-06 / 2009-07)– New design
– Portlet
– Can build Debian packages to make it easer to install the software
– Message of the day option
– 'Shopping basket' for publications and export of basket to a range of formats
Version 1.5 (2009-10)
– Advanced search
– MLIA implemented, based on Google Translate
– Datasets visible in portal
– Download statistics visible in portal
– 'Contextual' RSS for each page or search
Version 2.0 (2009-11 / 2009-12)
– Final improvements
EO portal
• DemoEO portal
• Available implementations:
– DSpace• various flavors
– Eprints.org• LSE• implemented by several partners
– Fedora• solution developped by UCLouvain• generic for all Fedora installations
DIDL/MODS implementation
OKvisible through the portal
EUR, KULeuven, LSE, ULB, UM, UvT
6
Good progress
tests are conclusive
tests are almost conclusive
Columbia University, Kiel
EUI, Sciences Po, UCL, Dauphine, UCD, Oxford, UCLouvain, CERGE, Carlos III
11
Progress Toulouse 1
No information Warwick 1
To start Monash, Geneva, Konstanz
3
DIDL/MODS implementation
Admin file implementation
OK - visible in portal EUR, LSE, ULB, UM, UvT
5
OK – not visible in portal
Kiel, UCD, Toulouse 3
Progress CERGE, Carlos III, KULeuven, Oxford, Columbia, EUI
6
No information Dauphine, UCL, Warwick, Sciences Po
4
To start Monash, Geneva, Konstanz, UCLouvain
4
• Allow for harvesting and integration of NEEO enriched metadata into local IR
• Integration with Copyright Knowledge Bank
• « Version signposting » tool / Cover page
Other local adaptations
• Technical guidelines version 1.3– MD5 encryption of IP addresses / privacy IPR– Robot filtering based on regular expressions– List of 50 regular expressions, based on input from:
• IR logs of Univ. Of Minho• IR logs of LSE
• Implementation:– DSpace: ULB
• alpha implementation
– Generic solution for Web logs: LSE (Tim Green)• generic solution based on Apache web logs
– Arno: UvT
• Status of implementation– Core: <30/4– non-core: now
Usage metadata
• Registering your usage metadata repository– Technical guidelines 1.4 – Annex 3– Revised Admin file template
• Usage metadata database in EO Gateway– Prototype planned for 31/8
• LogEC– Reuse LogEC data for RePEc publications ?
Usage metadata
• Meresco supports full-text indexing and searching
• Technique of deferred indexing
• New version of Meresco software has been delivered to UvT
• Protoype ready by 31/8
• OCR?• FT index of all OA RePEc publications?
Full-text search engine
• Meresco supports RSS service• Every search = RSS feed (CQL query)• RSS item
– title– abstract– permalink (to full record presentation of the publication
in the EO portal)• Todo
– title and description of feed– maximum number of records per feed– specs + impl EO portal
• Ready 31/8
RSS
RSS
RSS
• JEL– ready
• References– painful– prototype 31/8
• Integration into EO Gateway– concepts are clear– prototype 31/8
• reuse CitEC information for RePEc publications?
Enrichment
• Still waiting for integration into EconPapers
• EO Repec archive ready 31/8
EO RePEc archive
• CACAO 1 solution not integratable into EO portal• CACAO 2 ? TEXTEC / EXTRAKT ?• 31/8: solution based on Google Translate• In all solutions:
• Query translated into 3 other languages + expanded (based on dictionaries)
• All queries sent to our Meresco indexes• Multiple result sets• Presentation issue in EO portal: need more specs
• If analysis/choice for GT, CACAO, EXTRAKT conclusive: implement after 31/8 ?
MLIA
Solution Cost Control of dict
Google Translate 0 0
CACAO 9.000€ through CACAO
TEXTEC/EXTRAKT ? full control
• JTrac system
• Moved all issues from Google Docs document to JTrac
• Currently: 101 issues
• Public access to JTrac
Issue tracking
Issue tracking
Issue tracking
• Prototype implementations for all subtasks by 31/8
• Optimization by end of project
• WP5 deliverable 31/8
WP5 planning