Annif is an open source tool for automated subject indexing. Annif is developed at the National Library of Finland.
Annif utilises machine learning and language technology solutions. Annif is language independent and can be used with any indexing vocabulary, as long as the necessary language technology tools exist for the given language. Annif can be trained to better fit various use cases. The Finto AI service is based on Annif.
Annif development takes place on the Github platform and is open to public viewing and commenting. Annif can be downloaded from GitHub. It’s also available in the PyPI repository and as a Docker container. Annif has a web form for testing as well as a REST API for integration with other systems. We also provide downloadable, ready-to-use models for Annif that have been pre-trained with regard to the Finto AI service. An online tutorial on the use of Annif is also available for self-study.
In addition to the Finto AI service, Annif has also been utilised in e.g. ZBW - Leibniz Information Centre for Economics and the National Library of the Netherlands. Both institutions have also participated in the development of Annif.