INCEpTION - Interactive Concept Linking

Interactive Concept Linking

Example project

This project serves as an example for using Interactive Concept Linking supported by the recommendation functionalities of INCEpTION.

There are two documents in this example project:

  • pets1.tsv: A short annotated sample text about animals.
  • pets2.txt: A short annotated sample text about animals.

There is one Knowledge Bases included:

  • pets.ttl - includes some animals and instances

The following recommenders are configured in this project to give suggestions for named entity values and identifiers:

  • Token String Matcher - value
  • Token Sequence Classifier (OpenNLP NER) - value
  • Token String Matcher - identifier
  • Named Entity Linker - identifier

After importing the project, it should appear in Projects. A user can view the Documents, Knowledge Bases and Recommenders by clicking on the respective tabs. projects_settings

Please go to the Annotation Page and select pets1.txt to see example annotations. Then select the Named Entity layer. To annotate the identifier feature of a token, you can click on an existing annotation or mark any mention in a text and type in the identifier field to get suggestions. Grey annotations are recommendations by the system and can be accepted by clicking once or rejected by clicking twice. Detailed descriptions for using Interactive Concept Linking can be found in the INCEpTION User Guide. pets2.txt is an unannotated sample text that can be used as a playground.

annotation_page annotation_page

Download example project