hay / BOLD

Knowledge Graph Exploration and Analysis platform

BOLD ⋮

Knowledge Graph Exploration and Analysis platform

Features • Installation •

Features

TODO

Installation

The BOLD platform depends on postgresql and stardog databases for knowledge graph storage. In the following steps we discuss their setup as well as necessary steps to get BOLD up and running.

Database Setup

If you have already postgresql and stardog databases installed, you can skip this step. In this step we discuss their setup as docker containers.

• Install Docker and Docker Compose
• Build the Docker images: docker-compose build
• Start the database services: docker-compose up

Docker setup

If you want to install BOLD for development purposes, we refer you to the next section. In this step we describe steps on how to run BOLD as a docker container.'

• Build docker images: docker-compose -f docker-compose.full.yml build
• Start BOLD and the relevant services: docker-compose -f docker-compose.full.yml up
• Open the BOLD web interface: http://127.0.0.1:8000/

Development setup

For development purposes, you need to have a working python and rust installation.

• Install Rust and Python
• Build BOLD cli tools: make release-tools
• Install Poetry
• Install necessary dependencies: poetry install
• Run worker application: make start_worker
• Run backend server: make start_backend
• Open the BOLD web interface: http://localhost:8000/

Hybrid Setup

Use hybrid set up if you wnat to build the newest docker image of BOLD yourself. In this step we describe steps on how to run BOLD as a docker container.'

• Build docker images: docker-compose -f docker-compose.yml -f docker-compose.standalone.yml build
• Start BOLD and the relevant services: docker-compose -f docker-compose.yml -f docker-compose.standalone.yml up
• Open the BOLD web interface: http://127.0.0.1:8000/

System Requirements

The server requirements are mostly bound by the Stardog database.

You can choose to not use the Stardog database, but you will not be able to import the full datasets (only external SPARQL endpoints are allowed).

• You must have twice the amount of storage your datasets require. (YAGO is 60Gb thus 120Gb)
• You must allocate at least 2 cores for the server.
• Memory requirements are found below:

Number of Triples	Total System Memory
100 million	8G
1 billion	32G
10 billion	128G
25 billion	256G
50 billion	512G