Skip to content

Latest commit

 

History

History
30 lines (22 loc) · 1.42 KB

README.md

File metadata and controls

30 lines (22 loc) · 1.42 KB
Raw

rdf-converter

MeMAD metadata converter that transforms legacy metadata from INA and Yle into RDF using the MeMAD and EBU Core ontologies. MeMAD Ontology: http://data.memad.eu/ontology

Prerequisites

All the scripts are written in Python3. To run these scripts, the following libraries should be installed beforehand: pandas, tqdm and rdflib. These dependecies can be installed using pip.

Running the script

usage: python3 [ld|pa|yle]_converter.py [-h] [-p PATH] [-o OUTPUT] [-f FLOW_MAPPING] [-k]

MeMAD Converter

optional arguments:
  -h, --help            show this help message and exit
  -p PATH, --path PATH  Specify the path for the dataset (or datasets) to process
  -o OUTPUT, --output OUTPUT
                        Specify the path to which the TTL output would be
                        written.
  -f FLOW_MAPPING, --flow_mapping FLOW_MAPPING
                        Specify the path to a file containing the mapping
                        between filenames and their Flow identifier.
  -k, --keep_mappings   add this flag to generate CSV files for mapping
                        Programs to their URIs

!! For the subtitles mapping to work on INA's data, the Legal Deposit graph should be generated first (with the -k parameter) to guarantee that every subtitle line can be matched to its original program.