https://github.com/deezer/character_embeddings_qa
Repository for the paper Improving Quotation Attribution with Fictional Character Embeddings, Gaspard Michel, Elena V. Epure, Romain Hennequin, Christophe Cerisara, EMNLP 2024 (Findings)
https://github.com/deezer/character_embeddings_qa
Last synced: 9 months ago
JSON representation
Repository for the paper Improving Quotation Attribution with Fictional Character Embeddings, Gaspard Michel, Elena V. Epure, Romain Hennequin, Christophe Cerisara, EMNLP 2024 (Findings)
- Host: GitHub
- URL: https://github.com/deezer/character_embeddings_qa
- Owner: deezer
- Created: 2024-10-03T16:04:14.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-03-20T23:27:30.000Z (10 months ago)
- Last Synced: 2025-03-21T00:26:24.390Z (10 months ago)
- Language: Python
- Size: 50.9 MB
- Stars: 1
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Improving Quotation Attribution with Fictional Character Embeddings
This is the official repository for the EMNLP 2024 (findings) paper ["Improving Quotation Attribution with Fictional Character Embeddings"](https://aclanthology.org/2024.findings-emnlp.744.pdf). We train [LUAR models](https://aclanthology.org/2021.emnlp-main.70/) on Drama plays to distinguish utterances of fictional characters, and use the resulting models to derive character representations that are further injected in a [Quotation Attribution](https://aclanthology.org/2023.acl-short.64/) model to improve the accuracy on unseen literary works.
This repository contains two subfolders:
- **UAR** that includes all the code and data used to train LUAR on drama plays.
- **quotation_attribution**, a [clone](https://github.com/Priya22/speaker-attribution-acl2023) from [BookNLP+](https://aclanthology.org/2023.acl-short.64.pdf) where we modify the original quotation attribution model. This folder contains all code to train and reproduce our quotation attribution experiments.
# Installation
Run the following commands to create an environment and install all the required packages:
```bash
python3 -m venv charemb
. ./charemb/bin/activate
pip3 install -U pip
pip3 install -r requirements.txt
```
# Running the code
Each folder has its own README file, with instructions to run code.