CSV2KG

This repository contains a CLI that can be used to convert CSV files into "Semantic Knowledge". It can link the raw cell values to DBPedia entities, infer the probable class of a column and draw relations between the entities.

Usage

Usage: annotate.py [OPTIONS]

Options:
  --input_file TEXT   The CSV file to annotate
  --target_file TEXT  The target cells/columns/relations
  --output_dir TEXT   Where to store (temporary) results
  --help              Show this message and exit.

Example

python3 annotate.py --input_file=data/example_input.csv --target_file=data/example_targets.csv --output_dir=output

Input Data

Y. Khamis,Al-Wasl,Dubai,1982-07-23
W. Foulke,Chelsea F.C.,Shropshire,1874-04-12
T. Revill,Stoke City F.C.,England,1892-5-9

Targets

cell,2,6
cell,2,7
cell,2,8
cell,2,9
column,0,0
column,1,1
column,2,2
property,0,1

Output

Going to annotate the following file:
             0                       1           2           3
0    Y. Khamis                 Al-Wasl       Dubai  1982-07-23
1    W. Foulke            Chelsea F.C.  Shropshire  1874-04-12
2    T. Revill         Stoke City F.C.     England    1892-5-9
3  S. McMillan            Chelsea F.C.   Leicester  1896-09-17
4     S. James  KK Budućnost Podgorica    Brooklyn  1983-09-10

References

• Paper
• Poster
• Presentation

Written and maintained by Gilles Vandewiele and Bram Steenwinckel