LociSimiles¶

A Python package for extracting intertextualities in Latin literature using pre-trained language models.

LociSimiles enables researchers to detect textual reuse, quotations, and allusions between Latin texts, from verbatim citations to subtle paraphrases and thematic echoes.

Quick Start¶

Installation¶

pip install locisimiles

Or install with development dependencies:

pip install "locisimiles[dev]"

Basic Usage¶

from locisimiles import Document, RetrievalPipeline

# Load your documents
source = Document.from_csv("source_texts.csv")
target = Document.from_csv("target_texts.csv")

# Create retrieval pipeline
pipeline = RetrievalPipeline()

# Find similar passages
results = pipeline.retrieve(source, target, top_k=5)

Documentation¶

Getting Started - Installation and first steps
Examples - Working examples and tutorials
CLI Reference - Command-line interface documentation
API Reference - Complete API documentation
Development - Contributing and development setup

Authors¶

Julian Schelb - University of Konstanz
Michael Wittweiler - University of Zurich

Citation¶

If you use LociSimiles in your research, please cite our paper:

@article{schelb2026locisimiles,
  title={Loci Similes: A Benchmark for Extracting Intertextualities in Latin Literature},
  author={Schelb, Julian and Wittweiler, Michael and Revellio, Marie and Feichtinger, Barbara and Spitz, Andreas},
  journal={arXiv preprint arXiv:2601.07533},
  year={2026}
}

License¶

This project is licensed under the MIT License.