COLX 523 - Advanced Corpus Linguistics

Group Repository: David Kang, Daoming Liu, Jacob Nadal, Nicole Lopez

Project Overview

This is a fork of a repoistory dedicated to a group project for COLX 523 - Advanced Corpus Linguistics. The project involves building a corpus with annotation and a web interface, leveraging internet text sources to collect a sizable corpus (~1 million words).

Repository Structure

Movie-Review-Corpus-Annotation/
├── data/                # Raw and processed data files
├── documentation/       # Project-related documentation, meeting notes, and reports
├── src/                 # Source code for data collection, processing, and annotation
├── web_app/             # Contains the code for the interactive interface:
│     ├── frontend/      # The React application code for the user interface.
│     └── backend/       # The FastAPI application code for serving the API and static files  
├── .gitignore           # Files and directories to be ignored by Git
├── .dockerignore/       # Files and directories to be ignored by docker
├── instructions_movie-reviews.md   # Instructions for building and running the Docker image
├── Dockerfile/          # Dockerfile used to build a Docker image containing the entire project 
└── README.md            # Repository overview and guidelines

To create and run Docker image:

Please refer to intructions_movie-reviews.md in root repo.

Important File Locations (updated for sprint 3)

Annotation Data

• Human Annotations: Stored in Excel files under: src/analysis/modified_corpus_batches/xlsx/
• GPT annotations: Stored in json files under: src/analysis/modified_corpus_batches/json
• Code for annotation & evaluation: src/analysis/gpt and src/analysis/helper_methods

Interannotator agreement study and Plan for the interface: Both updated under documentation