Face to face meeting on VLO development January 2018

• What?
  • General VLO development progress, process and planning and meeting
• Who?
  • 16 January: Wolfgang, Matej, Menzo, Twan, Dieter
  • 17 - 18 January: Wolfgang, Menzo, Dieter, Thomas, Twan
• When?
  • 16 - 18 January 2018 (on basis of ​doodle)
  • Possibly 'satellite' in-depth dev knowledge exchange and/or pair programming on 15 and 19 January
• Where?
  • Utrecht and Nijmegen (see schedule)

Programme

Availability

Please fix/amend as needed:

Who	Mon 15	Tue 16	Wed 17	Thur 18	Fri 19
Dieter	Afternoon?	All day (?)	All day	All day	?
Matej	Arrival: ~ 11:00	All day	Departure ~ 19:00	-	-
Menzo	Will be in Nijmegen for a range of meetings	All day (if need be)	All day	All day (if need be)	Afternoon (if need be)
Wolfgang		Arrival: ~11:00	All day	All day	Departure: ~17:00
Thomas		Arrival: 17:00	All day	Departure: ~14:30
Twan	Afternoon	All day	All day	All day	All day

Schedule

Day	Location	Start	End	What?
Mon 15	Nijmegen(office)	14:00	17:30	Pre-meetings
Tue 16	Utrecht	11:15	17:00	Meeting: curation/quality
Wed 17	Nijmegen (sky lounge)	09:00	17:00	Meeting: VLO
Thu 18	Nijmegen (office)	09:00	17:00	Hands-on + discussions
Fri 19	Nijmegen (office)	09:00	16:00	Hands-on

Location details:

• Utrecht: ​Utrecht University
  • ​Drift 25 (enter via library)
    • Room 2.02
• Nijmegen: ​Radboud University
  • ​Erasmus Building
    • Office: room 8.06/8.07
    • Sky lounge: room 20.26 (​sky lounge on the 20th floor)

Agenda

See ​Google doc for agenda and notes

Documents

• ​Agenda and notes (see this page's attachments for export)
• ​Milestones
• ​Proposal for value mapping requirements and implementation

Notes

Aggregated action points

This week/ASAP

• {Menzo, Twan, Wolfgang} Implement compilation module prototype for curation workflow (CLAVAS + CSV ===Stylesheets===> uniform map)
• {Twan} Share curation workflow drawing source
  • See ​https://github.com/clarin-eric/VLO-mapping/wiki
• {Matej, Menzo, ...} prepare vocabularies/mappings for the curation workflow
  • See ​VLO-mapping fork (value-maps directory)
• {Matej} share SKOS evaluation spreadsheet

• {Twan} Move issue ​https://github.com/clarin-eric/VLO/issues/46 to VLO-mapping
• {GitHub issue/Twan/Wolfgang} Create github issue for integrating url checker (results) into the VLO (providing input to ‘can I access it’ -see #...)
  • See ​comment on #104

Vienna workshop

• Resourcetype: collection vs. corpus discussion
  • Make them synonyms in solr-configuration
  • Corpus narrower term of collection -> hierarchical facet
  • “Corpus” false positive Corpus Vitraeum
• Think about result sorting feature and/or discuss with Jakob

Soon

• Curation
  • Resource type facet: see if OTA values (154 distinct values) can be dealt with through concept mapping (blacklist?)  and/or value mapping
  • Find use cases for value mapping where context would allow for better mapping
    • For example ‘type’ field in olac
  • Look for corpora/collections flooded with individual item, aka forest-trees problem aka granularity problem
  • Olac-linguistic-type map to resource type rather than genre -> VLO-mapping
    • But also look into other cases where genre mapping might be justified
    • Make GitHub issue?

• Documentation, design etc
  • {??} Write-up/problem statement for value mapping
  • {??} Design and/or prototype for URL checker (based on curation module logic?)
  • {??} Make an updated functional/technical design for metadata quality dashboard

• Implementation / technical
  • {Wolfgang} Try to replace logic to get external VloConfig? in curation module
    • Ideal solution would be to ‘harvest’ it from running production VLO
  • {Menzo} integrate curation tool into the harvester
  • {Menzo} Deploy harvest viewer
  • {Menzo} export mapping from directory name to endpoint, center, national project
    • {Twan} Add field to VLO for OAI endpoint (‘data provider’)
      • {GitHub issue} ​https://github.com/clarin-eric/VLO/issues/131
  • {Twan} Make a GitHub issue for Java 9 support in the VLO
    • {GitHub issue} ​https://github.com/clarin-eric/VLO/issues/132
  • {???} Make a common library/tool for url checker (see todos of 16 January)
  • {Twan} EDM-CMDI conversion (fix or make issue):
    • resolve library of congress subject heading codes to preflabel
    • Also getty for resource type e.g. ​http://vocab.getty.edu/aat/300026656
    • {GitHub issue} ​https://github.com/clarin-eric/metadata-conversion/issues/3
  • {GitHub issue} Look into weird querying behaviour “german OR dutch” vs “(german OR dutch)”
    • ​https://github.com/clarin-eric/VLO/issues/129
    • Attempt at solution has been applied to alpha (see issue comments), to be tested...
    • Solution applied to production, to be consolidated with deployment of VLO 4.3.3
  • {GitHub issue} Similar record folding
    • ​https://github.com/clarin-eric/VLO/issues/113
  • {Twan} Find info on “copy/paste problem” in old solr logs (e.g. ‘Talk of Norway’)
    • Note: Solr logs indicate that these searches (around 4-6 December) did result in hits; perhaps a problem in the front-end (in particular ​#118) prevented them from appearing

Later

• Documentation, design etc
  • Investigate the creation of landing pages for specific languages (manually curated + links to VLO)

• Implementation/technical
  • {Curation module} Field-weighting in curation module
  • {Curation module} investigate curation score of BBAW
  • {Curation module} use VLO importer as a library to share the mapping and processing logic rather than have it duplicated
  • {Menzo, ?} Adapt harvester viewer to fully fledged MD quality dashboard
  • {Twan} VLO: icons per record or collection
    • See ​https://github.com/clarin-eric/VLO/issues/54
  • {Dieter} Add VLO and FCS search to drupal search widget on www.clarin.eu (ask Hendrik?)
  • {Twan} VLO: see if it is possible to suggest discriminative additional search terms wrt query results (make GitHub issue)