web-sand

SAND

Table of Contents

• Introduction
• Installation

Introduction

SAND is an application to annotate semantic descriptions of tables and (optionally) linked records in tables to a target knowledge graph, then it can automatically export the table data to RDF, JSON-LD, etc. It also does basic data cleaning automatically based on the annotated semantic descriptions. SAND is designed to be customizable: you can plug in a new semantic modeling algorithm (which generates a semantic description automatically) or different knowledge graphs as long as you have a suitable plugin implemented SAND's plugin interface.

Moreover, SAND offers an internal KG browsing and table filtering so you can interactively browsing and modeling your tables.

For a demo, please see: our demo paper, demo video.

Installation

Install from pip: pip install -U web-sand

Usage

• Initialize database: sand init -d <dbfile>. For example: sand init -d ./data/sand.db
• Start the webserver: sand start -d <dbfile> --externaldb <folder_of_ent_and_ont_db>
• Open the URL: http://localhost:5524

For example, checkout server.sh

Development

• cd to www
• Install yarn and yalc
• Install dependencies: yarn install
• Start development server: yarn start. Then, access development server at: http://127.0.0.1:3000.
• Build production files: yarn build
• Build library files and publish to private index (only if you are released sand as a library): yarn build:lib && yalc public --private