vialab/citation-galaxies

Citation-Galaxies

Bibliometric experts use statistical analysis of natural language texts to provide a quantitative view of academic literature. Bibliometric techniques, such as citation analysis, are often used by funding agencies to understand and evaluate the publishing results of grant money distribution. While this field of scholarship in general, and citation analysis in particular, have been studied extensively, there still remains a lack of tools and techniques for evaluating the surrounding natural language context of individual citations on the network scale. The goal of this project is to aid in the identification of language patterns as they occur across a large corpus of academic texts, allow users to create and edit rule sets for evaluation, and then use vector space models to suggest expansions of the user defined natural language rules. Although originally developed for citation analysis, the tool can be customized for any number of categories and rules that apply to a given domain problem.

Getting Started

This project was built as two different servers:

• NodeJS primarily for frontend
• Python (Flask) primarily for backend data processing

Please ensure that you have docker installed on your system. Both projects are setup to run in docker containers either separately, or using docker compose. By default, these projects run on port 5432 for the frontend, and port 5000 for the backend.

Running with Docker

Run both frontend and backend sub-folders as seperate images.

Frontend:

• Navigate to /frontend
• Build the docker image using docker build -t cg-frontend .
• Run the container with docker run cg-frontend

Backend:

• Navigate to /backend
• Build the docker image using docker build -t cg-backend .
• Run the container with docker run cg-backend

By default, the project (frontend) should be accessible through localhost:5432

Running Locally

A connection to the database must be made in order for the proper functioning of this project. Both frontend and backend will require a connection. Please contact the collaborators of this project for access to the database.

Backend:

• Create a file in the /backend sub-folder with the name database.ini
• Paste the following into the database.ini file, edit the values as appropriate and save

[postgresql]
host=<host>
database=citationdb
user=citationdb
password=<password>

• You may now run this project as you would for a regular FLASK project
• Example:

#!/bin/bash
echo "Starting python server.."
source venv/bin/activate
export FLASK_APP=server.py
export DEPLOY_ENV=TEST
flask run --port=5000
echo "Python listening on http://localhost:5000"

Frontend:

• First navigate to the /frontend sub-folder in your terminal
• Run the command npm install
• Running the frontend project will require environment variables be set available to the nodejs project, and so we would recommend you to write your own bash script to handle this (or whatever is available for your respective system)
• Example:

#!/bin/bash
DATABASE_URL="postgresql://citationdb:<password>@<host>:<port>/citationdb" DEPLOY_ENV="PROD" node index.js

Alternatively, you may edit the Dockerfiles as you see fit for your setup. However, please DO NOT commit/push any changes to the Dockerfiles unless you know what you are doing.

Deployment

Deployment for this project has been automated, and so please be aware that pushes to this repository will automatically build, run, and deploy to the VIALAB production servers at [***] Deployment will automatically handle database connections, as well as mount volumes to hold the larger files not part of the github repository (citation caching and word2vec models) through Kubernetes.

In order to ensure the appropriate database connections are made, committed versions of this project should always have ENV DEPLOY_ENV PROD in line 6 of the Dockerfile for the /backend sub folder.

Built With

• https://nodejs.org/en/ - The web framework / packaging system used
• https://getbootstrap.com/ - Front-end component library
• D3 - Visualization Library
• https://www.docker.com/ - Container / Dependency management
• http://flask.pocoo.org/ - The web framework used (PYTHON 2.7)

Authors

• Adam Bradley, PhD. - Research Associate
• Christopher Collins, PhD. - Research Supervisor
• Kevin Desousa - Research Assistant
• Victor (Jay) Sawal, BSc. - Software Developer

License

This research was conducted as part of the CO.SHS project (co-shs.ca) and has received financial support from the Canada Foundation for Innovation (Cyberinfrastructure Initiative – Challenge 1 – First competition).