PyTerrier Artifacts

PyTerrier Artifacts provide a powerful way to share resources, such as indexes, cached results, and more. Re-using one another’s artifacts is a great way to help achieve green (i.e., sustainable) research.

Artifacts are ready-to-use in your experiments, since they expose their functionality with Transformers. For instance, once you load a TerrierIndex artifact, you can use its bm25() method to build a transformer that retrieves from the index using BM25.

Contents

• How-To Guides
• Listing
• API Reference

Acknowledgements: The design of the Artifact API was described in the following paper:

Citation

MacAvaney. Artifact Sharing for Information Retrieval Research. SIGIR 2025. [link]

@inproceedings{DBLP:conf/sigir/MacAvaney25,
  author       = {Sean MacAvaney},
  editor       = {Nicola Ferro and
                  Maria Maistro and
                  Gabriella Pasi and
                  Omar Alonso and
                  Andrew Trotman and
                  Suzan Verberne},
  title        = {Artifact Sharing for Information Retrieval Research},
  booktitle    = {Proceedings of the 48th International {ACM} {SIGIR} Conference on
                  Research and Development in Information Retrieval, {SIGIR} 2025, Padua,
                  Italy, July 13-18, 2025},
  pages        = {3974--3979},
  publisher    = {{ACM}},
  year         = {2025},
  url          = {https://doi.org/10.1145/3726302.3730147},
  doi          = {10.1145/3726302.3730147},
  timestamp    = {Sat, 09 Aug 2025 12:14:08 +0200},
  biburl       = {https://dblp.org/rec/conf/sigir/MacAvaney25.bib},
  bibsource    = {dblp computer science bibliography, https://dblp.org}
}