ARAGORN-ranker

The ARAGORN ranker used by the ARAGORN ARA, which takes a TRAPI 1.5 message containing answers, and calculates numerical scores for each answer.

This is a upgraded port of code from robokop-messenger to perform omnicorp overlay and score for answer ranking.

Omicorp overlay attaches literature co-occurrence support graphs to each result, and the score operation calculates the overall score of the result. Together, the support graphs and the score are included as an analysis of of the result.

ARAGORN-ranker uses the omnicorp database to retrieve ontologies and perform literature co-occurrence calculations. Please see the link below that references that codebase.

Demonstration

A live version of the API can be found here.

Related Source Code

Below you will find references that detail the standards, web services and supporting tools that are part of ARAGORN.

• ARAGORN
• Omnicorp

Installation

To run the web server directly:

Create a virtual Environment and activate.

cd <aragorn-ranker root>

python<version> -m venv venv
source venv/bin/activate

Install dependencies

pip install -r requirements.txt

Run Script

cd <aragorn-ranker root>

./main.sh

DOCKER

Or build an image and run it.

cd <aragorn-ranker root>

docker build --tag <image_tag> .

Then start the container

docker run --name aragorn-ranker -p 8080:4868 aragorn-test

Kubernetes configurations

Kubernetes configurations and helm charts for this project can be found at:

https://github.com/helxplatform/translator-devops/helm/aragorn-ranker