Data Mining and Machine Learning Algorithm Annotator

Overview

The ML Algorithms Annotator is a Django application which enables users to annotate and query ML algorithms. The algorithms information is based on an annotation schema and it includes entities like "Document", "Optimization Problem", "Complexity", etc. Additionally, the user can query the graph database containing multiple PoC algorithms like Linear Regreession, SVC, KNN, etc.

The development process is described in detail in Chapter 4 in my MSc thesis (p. 39).

Installation

Python dependencies

1. Install with conda (using environment.yml) - TBD
   conda env create -f environment.yml
2. Install with conda (using requirements.txt)
   conda create --name <env_name> --file requirements.txt
conda activate <env_name>
3. Install with venv
   python3 -m venv env_name
source ee/bin/activate
python3 -m pip install -r requirements.txt
4. Install package - TBD
   pip install .

Node dependencies

1. Install node.js
2. cd into the frontend folder
   cd web-annotator/controller/frontend
3. Install npm
   npm install -g npm
4. Install dependencies
   npm i
5. Compile the front-end
   npm run build - production
   npm run dev - development

Neo4j setup - TBD

Run the application

1. cd to the controller folder
   cd web-annotator/controller
2. Run the server
   python manage.py runserver

Support

lidija.jovanovska@outlook.com