cbobed/swkrimpsim

Project for the use of the Krimp algorithm for the measure of similarity between RDF datasets

#SWKrimpSim: Structural similarity of RDF graphs using frequent pattern mining #

Content of the repository

Source code of the approach for a structural similarity of RDF graphs based on frequent patterns, and its related experiments. In this project, it is included the code for the proposal to evaluate the evolution of a dataset currently submitted to SWJ.

• The SWPattern folder contains the source code for the conversion of RDF datasets and updates to transaction, and their comparison and evaluation using codetables.

• The Scripts folder contains the scripts used to retrieve code tables from Vreeken's implementations of KRIMP and SLIM.

• The Slim archive contains our modification of Vreeken's SLIM implementation to be able to handle very large number of items (by removing a hard coded limit).

• The PythonCode folder contains the code used for the data analysis relative to our experiments.

Citation

To cite the structural similarity measure related to this approach:

@inproceedings{Maillot2018, 
    author = {Pierre Maillot and Carlos Bobed}, 
    title = {Measuring Structural Similarity Between RDF Graphs}, 
    booktitle = {Proc. of 33rd ACM/SIGAPP Symposium On Applied Computing (SAC), SWA track}, 
    publisher={ACM}, 
    pages = {1960--1967}, 
    month = {April}, 
    year = {2018}, 
}

To cite the assessment of the structural evolution of RDF (Knowledge) graphs:

@article{Bobed2019,
    author={Carlos Bobed and Pierre Maillot and Peggy Cellier and S\'ebastien Ferr\'e},
    journal={Semantic Web},
    month={January},
    number={1},
    pages={831--853},
    publisher={IOS Press, ISSN 1570-0844},
    title={Data-driven Assessment of Structural Evolution of RDF Graphs},
    year={2020}
}