Annif is a tool for automated subject indexing. To work, it needs a controlled vocabulary (subject headings, thesaurus or classification) and existing metadata - Annif can then be used to assign subjects for new documents. This tool is built upon a combination of existing natural language processing and machine learning tools including Maui, Omikuji, fastText and Gensim. It is designed to be multilingual and it can support any subject vocabulary (in SKOS or a simple TSV format). It can be used either via a command-line interface or a microservice-style REST API. You can read more about using Annif in the GitHub Wiki, the API is described shortly below.
Annif API moving towards production
Annif has a demo API under Api.annif.org. The API can also be used via the form at annif.org. The main detalis are briefly:
The API uses version 0.46 of Annif
We offer a new neural network based backend for projects (nn_ensemble), trained with a broad training set from Finna
The API now uses a recent version of the General Finnish Ontology with place names included (version 2020.1.Diotima)
The API offers the following projects yso-fi, yso-sv, yso-en (implemented with nn_ensemble), wikidata-en and hogwarts
The API works over a secure HTTPS connection.
In the near future, after some further research and development, we’ll aim to offer a learn method via the API, so that human-corrected results could be used to teach Annif and improve results. Also, we're taking Annif into production by offering the API under the name Finto AI - stay tuned!
