Generic UIMA pipeline

Simple UIMA pipeline wrap as a console application annotating Sentences, Items and Events. The last ones are easily customisable by updating the current list of terms.

Generated annotations:

• Sentences
  • Annotation
    • de.tudarmstadt.ukp.dkpro.core.api.segmentation.type.Sentence
  • Annotator
    • OpenNlpSegmenter: https://dkpro.github.io/dkpro-core/releases/1.9.0/docs/component-reference.html#engine-OpenNlpSegmenter
• Items
  • Annotation
    • genericpipeline.Item
      • property: 'value'(String)
  • Annotator
    • UIMA RUTA MARKTABLE
    • Resources
      • word table taken from: ruta_resources/item_table.csv
        • The first column contains a list of unique items to be tagged in the text.
        • The second column value is assigned to the property 'value'
• Events
  • Annotation
    • genericpipeline.Event
  • Annotator:
    • UIMA RUTA MARKFAST
    • Resources
      • word list taken from: ruta_resources/event_list.txt

Input

• Directory containing text files

Output

• XMI Files

Usage

1. Download and extract the application: https://github.com/ewoij/generic-uima-pipeline/releases/tag/v1.0.0
2. Update if necessary the following files with your own domain terms:
   • ruta_resources/item_table.csv
   • ruta_resources/event_list.txt
3. Run from the current directory

   java -Xms4000m -jar generic-uima-pipeline-1.0.0-standalone.jar <input folder containing text files> <output folder>
   

Resources

• item_table.csv was built from a subset of NCI-Thesaurus