Spaces:
Running
Running
| cff-version: 1.2.0 | |
| message: "If you use Paraphrasis in your work, please cite both the software (below) and the article listed under 'references'." | |
| title: "Paraphrasis" | |
| abstract: >- | |
| Paraphrasis is an online research tool that makes a full semantic corpus | |
| analysis pipeline available to non-specialists: automatic segmentation of raw | |
| texts, unsupervised fine-tuning of an existing sentence-embedding model on | |
| user-provided sequences, and search for semantic correspondences between two | |
| corpora through semantic textual similarity. | |
| type: software | |
| authors: | |
| - given-names: "Théotime" | |
| family-names: "de la Selle" | |
| affiliation: "Institut des Sources Chrétiennes, HiSoMA, CNRS, Lyon, France" | |
| orcid: "https://orcid.org/0000-0002-3591-3294" | |
| email: "theotime.de-la-selle@cnrs.fr" | |
| version: "0.1.0" | |
| date-released: "2026-06-19" | |
| license: MIT | |
| repository-code: "https://huggingface.co/spaces/TdelaSelle/Paraphrasis" | |
| url: "https://tdelaselle.gitpages.huma-num.fr/paraphrasis/" | |
| keywords: | |
| - "text reuse detection" | |
| - "Information retrieval" | |
| - "semantic textual similarity" | |
| - "sentence embeddings" | |
| - "unsupervised representation learning" | |
| - "ancient languages" | |
| - "Latin" | |
| - "Ancient Greek" | |
| - "digital humanities" | |
| - "research software" | |
| references: | |
| - type: article | |
| title: "From Transcription to Semantic Corpus Analysis: Unsupervised Learning of Sentence Representations for Ancient Languages" | |
| authors: | |
| - given-names: "Théotime" | |
| family-names: "de la Selle" | |
| affiliation: "Institut des Sources Chrétiennes, HiSoMA, CNRS, Lyon, France" | |
| year: 2026 | |
| month: 7 | |
| url: "https://arxiv.org/abs/2607.24542" | |
| identifiers: | |
| - type: other | |
| value: "arXiv:2607.24542" | |
| description: "arXiv preprint identifier (cs.CL), v1, 27 July 2026" | |
| notes: "Preprint presenting the methods implemented in Paraphrasis. Update to the journal reference once published." | |