-
Currently, the `eval` command allows to evaluate the predictive performance of a backend for a single training and test split. Unfortunately, splitting multi-class multi-label data into training and t…
-
Create a list of models that we could potentially use to extract text from documents and suggest metadata. We will start with basic metadata like title, description, etc. and eventually move on to op…
-
## Description of the enhancement
Skosmos should display a nice API documentation such as https://api.finto.fi/doc/ by default at `/rest/v1/`. This can hopefully be generated from existing Swagger …
-
Hi @cbroschinski, thank you for sharing your work!
What is the license of [base_dump](https://github.com/cbroschinski/MALIS-21-Master_Thesis/tree/master/data/base_dump)? Can I reuse it for training…
-
TUPrints needs to know DDC classification of the publication. This is likely not to be automated and hardly any metadata to find, so add a field to the form.
It should be considered to add a genera…
-
Hi,
Maybe a framework like VueJS could be used instead of HTML + JS (JQuery + Bootstrap). I think the use case so far is perfect for a small app as Annif is right now.
But for #24 and onward, I …
-
Neben den freien Schlagworten `"keywords"` wäre es sinnvoll kontrollierte inhaltliche Vokabulare als Schlagworte jenseits der Fachdisziplinen (`"about"`) abzubilden.
-
We are currently using Gensim only for the basic TF-IDF backend. It should be almost trivial to create an LSI backend, it's just one extra LsiModel layer and a single parameter (number of dimensions).…
-
Sometimes, reconciliation queries can be invalid, for all sorts of reasons:
* their JSON structure does not fit the schema
* they contain references to objects that do not exist (for instance, inval…
-
In Automated subject indexing (and multi label classification in general) the distribution of assigned concepts often follows Zipf's law. In our experience this leads to algorithms having low precisio…
mo-fu updated
12 months ago