EccLexica : Christian Architectural Discourse Detection in Latin Texts

This repository contains a dictionary-based method for identifying architectural discourse in Late Antique and Medieval Latin texts. It is applied to a test corpus of Latin inscriptions.

The workflow combines a dictionary-based approach with an evaluation against GLiNER named-entity recognition models.

EccLexica is being developed as part of the ANR E-cclesia project.

---

Repository Structure

.
├── data/
│   └── dicts
│     └── auto_terms.csv
│     └── asso_terms.csv
│     └── mat_terms.csv
│   └── sample.csv
│
├── output/
│
├── inscr-inscr-preprocess.py
├── inscr-inscr-lemmatize.py
├── inscr-inscr-score.py
├── gliner-eval.py
└── README.md

---

Data

• Main dataset

The main dataset is not stored in this repository due to its size.

Please download EDCS_text_cleaned_2022-09-12.json from Zenodo:

https://zenodo.org/records/7072337

After downloading, place the file in the following location:

data/EDCS_text_cleaned_2022-09-12.json

• Sample dataset (data/sample.csv)
  • 100 inscriptions.
  • Used for testing and GLiNER evaluation.

---

Scripts

This repository includes three standalone command-line scripts for a split workflow:

1. inscr-preprocess.py — filter raw JSON inscription data by date and province.
2. inscr-lemmatize.py — lemmatize filtered texts with spaCy.
3. inscr-score.py — calculate architectural scores from lemmatized texts.

---

Requirements

• Python 3.8+
• pandas
• numpy
• spacy
• gliner
• Latin spaCy model: la_core_web_md

Install dependencies if needed:

pip install pandas numpy spacy
python -m spacy download la_core_web_md

---

1. Preprocess

This step loads a JSON file, filters inscriptions by the default date range and excluded provinces, and saves the filtered CSV.

python inscr-preprocess.py \
  --input data/EDCS_text_cleaned_2022-09-12.json \
  --output data/edcs_filtered_inscriptions.csv

Use --exclude-provinces to supply a comma-separated list of provinces to exclude without an interactive prompt:

python inscr-preprocess.py --exclude-provinces "Achaia,Aegyptus,Syria" --no-prompt

---

2. Lemmatize

This step reads the filtered CSV and adds a lemmatized_text column.

python inscr-lemmatize.py \
  --input data/edcs_filtered_inscriptions.csv \
  --output data/edcs_lemmatized_inscriptions.csv

---

3. Score

This step reads the lemmatized CSV, calculates architectural scores, and writes the scored output.

python inscr-score.py \
  --input data/edcs_lemmatized_inscriptions.csv \
  --output output/edcs_architectural_scores.csv \
  --high-output output/edcs_architectural_scores_gt50.csv

To disable creation of per-score-bin subset CSV files, use:

python inscr-score.py --no-subsets

---

Notes

• inscr-preprocess.py may prompt interactively for excluded provinces unless --exclude-provinces or --no-prompt is used.
• inscr-lemmatize.py uses spaCy and needs the model installed.
• inscr-score.py expects the lemmatized data to contain a lemmatized_text column.

gliner-eval.py

Evaluates GLiNER NER models against the dictionary-based architectural terms.

What it does:

• Loads a scored sample dataset from output/sample.csv.
• Runs multiple GLiNER models on:
  • Lemmatized text
  • Cleaned interpretive text
• Compares GLiNER-extracted entities with dictionary-based terms.

Evaluated labels:

• building/type
• building/part
• building/material

Metrics computed:

• Precision
• Recall
• F1 score
• Percentage and count of overlapping terms
• Number of inscriptions with at least one shared term

---

Notes

• Scores are heuristic and intended for comparative and exploratory analysis, not absolute classification.

---

Citations

Heřmánková, Petra. “EDCS”. Zenodo, September 12, 2022. https://doi.org/10.5281/zenodo.7072337.