softalias-rs

Softalias-rs is a service that uses Softalias-KG as a reconciliation service for similar tool mentions, which we have integrated with the Named Entity Recognition (NER) model Softcite trained in the biomedical domain to extract software mentions.

Demo: Avaliable here.

Authors: Esteban Gonzalez and Daniel Garijo

Requirements:

Softalias-rs has been tested in Unix operating systems.

• Python 3.8 - 3.11
• PIP
• Streamlit 1.24
• Softcite service 0.7.1.

Make sure you have deployed the Softalias-KG in a SPARQL endpoint (our demo service uses https://softalias.linkeddata.es/sparql)

To install streamlit, follow these instructions

To install a service of softcite, follow these instructions. An open softcite service can be found here

Install from GitHub

To run softalias-rs, please follow the next steps:

Clone this GitHub repository

git clone https://github.com/SoftwareUnderstanding/softalias-rs.git

Install the python libraries required.

cd softalias-rs
pip install -e .

To run the service, execute

streamlit run app.py

The service will be available in the port 8501.

Install through Docker

We provide a docker image for the service.

First, you have to build the docker image

docker build -t softalias-rs .

Then, you can run the service with the command

docker run -d softalias-rs

Contribute

Pull requests are welcome in the main branch.