Recent Releases of nederlab-pipeline

nederlab-pipeline - v0.9.11

Bugfix release and added extra check to guard against a certain corruption in the metadata database

- Groovy
Published by proycon over 5 years ago

nederlab-pipeline - v0.9.10

Bugfix release

- Groovy
Published by proycon over 5 years ago

nederlab-pipeline - v0.9.9

Further fixes in DBNL fix pipeline

- Groovy
Published by proycon over 5 years ago

nederlab-pipeline - v0.9.8

Incorporated extra validation step in fix pipeline

- Groovy
Published by proycon over 5 years ago

nederlab-pipeline - v0.9.7

Minor update for dbnl fix pipeline: Reverted unnecessary patch from previous release and ensure output files can't overwrite input

- Groovy
Published by proycon over 5 years ago

nederlab-pipeline - v0.9.6

Minor update for dbnl fix pipeline: added sync in an attempt to tackle an elusive issue

- Groovy
Published by proycon over 5 years ago

nederlab-pipeline - v0.9.5

Bugfix release in fix pipeline: Unassign unused metadata

- Groovy
Published by proycon over 5 years ago

nederlab-pipeline - v0.9.4

Bugfix release for fix pipeline: Acts can also be independent chapters

- Groovy
Published by proycon over 5 years ago

nederlab-pipeline - v0.9.3

Minor release update; better gz compression in fix pipeline

- Groovy
Published by proycon over 5 years ago

nederlab-pipeline - v0.9.2

Further bugfix release for dbnl fix pipeline

- Groovy
Published by proycon over 5 years ago

nederlab-pipeline - v0.9.1

bugfix release for dbnl fix pipeline

- Groovy
Published by proycon over 5 years ago

nederlab-pipeline - v0.9.0

Recent changes:

  • Implemented a script that fixes the DBNL FoLiA v2 documents as delivered in 2019. This scripts fixes the IDs and adds the necessary (sub)metadata. Discussed in Jira ticket: https://jira.socialhistoryservices.org/browse/TT-709

Older changes (2019):

  • enable ignore option for wikiente
  • implement support for language constrain in modernisation
  • added resources (migrated from inl/nederlab-linguistic-enrichment)
  • only do language identification on sentences!
  • simplifying the pipeline, do not run frog in batches anymore but one frog per file (at cost of init time and extra memory, but easier to handle potential errors)
  • replacing folialangid with colibri-lang, use --subcodes for colibri-lang
  • do language detection before tokenization

- Groovy
Published by proycon over 5 years ago

nederlab-pipeline - v0.8.0

This is a first release after the split from PICCL. Introduces a renewed Nederlab pipeline consisting of the following components (most can be enabled/disabled when needed): * Conversion from TEI to FoLiA through tei2folia (part of FoliA tools, this replaces the old scripts by Erik Tjong Kim Sang. * FoLiA v2 compatible, stricter and explicit validation and explicit provenance information. * Tokenisation through ucto * Language detection using folialangid (part of FoliA tools) * Linguistic enrichment through Frog (part-of-speech tagging, lemmatisation, named entity recognition) * Named Entity Linking (after another recognition stage) is handled by wikiente, which communicates with the DBPedia Spotlight service, and replaces the older foliaentity tool by Erwin Komen.

- Groovy
Published by proycon about 7 years ago