metadata

dataset_info:
  features:
    - name: vat
      dtype: string
    - name: pubid
      dtype: string
    - name: act_description
      dtype: string
    - name: company_name
      dtype: string
    - name: company_juridical_form
      dtype: string
    - name: address
      dtype: string
    - name: street
      dtype: string
    - name: zipcode
      dtype: string
    - name: city
      dtype: string
    - name: publication_date
      dtype: date32
    - name: publication_number
      dtype: string
    - name: publication_link
      dtype: string
    - name: text
      dtype: string
    - name: is_digital
      dtype: bool
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/year=*/month=*/day=*/*
license: apache-2.0
task_categories:
  - text-generation
language:
  - fr
  - nl
  - de
tags:
  - finance
  - legal
pretty_name: Belgian Journal (Moniteur Belge/Belgisch Staatsblad)
size_categories:
  - 100K<n<1M

Dataset Description

The most up to date data can be found here: https://huggingface.co/datasets/guust-franssens/belgian-journal Due to the different languages in Belgium, I decided to create three datasets belgian-journal/moniteur-belge/belgisch-staatsblad in order to make it more visible.

Dataset contains the metadata + the text of bylaw publications of Belgian companies on the Belgian Journal (Moniteur Belge/Belgisch Staatstblad). This data was collected by webscraping the Belgian Journal, for more info see: https://github.com/Guust-Franssens/belgian-journal.

Language(s) (NLP): French, Dutch and a small subset German (official languages of Belgium.)
License: apache-2.0