D-Square

1
D-Square

• Digital Databases and Digital Tools
• for
• WBD and WLD
• Folkert de Vriend
• 17-05-06

2
Outline

• Digitisation project (shortly)
• Plans and ideas for papers
• Data driven clustering
• Open Language Resources
• Cartography

3
People
CLST Lou Boves Henk van den Heuvel Folkert de
Vriend CLS Roeland van Hout Joep Kruijsen Jos
Swanenberg Polderland Theo van de Heuvel
4
WBD page -gt
5
(No Transcript)
6
Data conversion overview -gt
7
Editors / Management Users Analog Digital Digital Analog
Raw data FileM Pro
Edited data XML
Raw data
Questionnaires Nijmegen and Leuven
Questionnaires (chiefly) Meertens
(parts of) Vol. III MS-Word
Vol. III FileM Pro
Enriched data XML
(parts of) Vol. III MacWrite
Deel III MS-Word
Vol. III
Filing cards
Online DB WBD (Polderland)
Edited data
Vol. I II
Vol. III
Website WBD/WLD with tools for searching
and cartography
Specialized print editions (dialect atlas or
local dictionary)
SGV on CD (Polderland)
8
Web access
Taxonomic acces to data Search interface
9
Research ideas and plans
10
A Data driven clustering

• Human interpretation of patterns
• vs
• computational clustering based on distances.
• (lexical or phonetic)

11
(No Transcript)
12
(No Transcript)
13
B Open Language Resources

• Wikipedia style LR
• Digitisation not the end of the evolution of a
  LR
• Evolution of Web seems to be towards Social
  Computing
• Think of railroads -gt cars

14
Policing

• How to automate police activities regarding open
  (language) resources?
• Maybe distant entries/edits are more
  suspicious. When distant -gt notify police.

15
C Cartography

• Cartography as tool not just for illustration -gt
  Google Earth.
• Advantages
• Different views on the data.
• Easy to link different resources (also for end
  user)

16
Implementation

• Short term
• Paper on data driven clustering.
• Paper on cartography.
• Longer term
• Paper(s) on Open LR / Social Computing.