INCEpTION - Recommenders and Active Learning

Recommenders and Active Learning

Example project

This project is an example for using the recommendation and active learning functionalities of INCEpTION to annotate the part-of-speech tags.

There are two documents in this example project:

  • GUM_interview_chomsky.tsv: it is a well-annotated document chosen from the GUM Corpus
  • Noam_Chomsky.txt: it is a part of the Wikipedia article Noam Chomsky

Two following recommenders are configured in this project to give suggestions of layer POS:

  • Token String Matcher
  • Token Sequence Classifier(OpenNLP POS)

After importing the project, this project appears in Projects. A user can click on this project, then switch to Documents to view the uploaded documents, and switch to Recommenders to view the configured two recommenders:

projects_settings

After that, a user can switch to the annotation page and open the document Noam_Chomsky.txt. The recommendations are shown above the tokens. A user can find the detailed description for using active learning to do the annotations in the INCEpTION User Guide. By clicking Help, a user can view this user guide.

annotation_page


Download example project