Allows to grab your very own wording database with semantic relations

opensemanticapi with NodeJS

This project exists in two different languages:

Works with NodeJS, Restify, UnderscoreJS and Redis


Will allow you to create your own semantic wording database with redis. Otherwise there will be a open api to get related words by meaning. You could say, this implementation is a light version of the idea behind in combination with



The following examples where given after the system was collecting for about one hour only.

1. Example (http://localhost:8080/relations/ship):

  • Input: "ship"

  • Output: ["midshipmen", "aboard", "ships", "rating", "master", "served", "seaman", "sea", "officers", "santa", "sailing", "cadets", "able", "sail", "navigation", "lieutenant", "hms", "expected", "yahoo", "storm", "rated", "promotion", "maría", "lewis", "false", "era", "boys", "wind", "voyage", "volunteer", "servants", "required", "passing", "palos"]

2. Example (http://localhost:8080/relations/human):

  • Input: "human"

  • Output: ["humans", "evolution", "primates", "ago", "ado", "studies", "physiology", "bonobo"]

3. Example (http://localhost:8080/relations/dog):

  • Input: "dog"

  • Output: ["infant", "wildlife", "offspring", "mother", "future", "southwest", "koalas", "conflict", "animals", "aitken", "wolf", "urban", "rehabilitation", "pet", "perspective", "nursing", "mexico", "evolutionary", "weaning", "ticket", "texas", "speech", "special", "retrospective", "primate", "holtcamp", "fund", "enough", "domestic", "cost", "arizona", "210–217", "variety", "trivers", "trauma", "terms", "sprawl", "southwestern", "sense", "river", "received", "questions", "point", "perhaps", "parent", "otter", "makes", "little", "less", "himself", "gray", "gorilla", "frequently"]

Installation Guide

  • install npm and node if you have not already ( or
  • install / start your redis server ( on a disk with several free GB

  • clone this repo "git clone"

  • change config if needed in "/config.js"
  • NOTE: depending on what you want (scraper / cronjob to run or http api server, set 'http_server' to true or false)
  • open the repository folder in your console
  • enter "npm install", it will install all dependencies automatically
  • start the node server with "node app.js"
  • now it should print what it is collecting or what http route is requested
  • the longer it collects data the better the results should be
  • now you can access the relations through your browser like http://localhost:8080/relations/database or direct by accessing your redis server

To Do

  • improve performance of script
  • allow certain configurations (half done)
  • try to use with its structure for more compatibility and understanding
  • write tests
  • connect to travis-ci
  • better folder structure
  • write version using level db or similar

Related Links

To support this project, please consider to donate.

This software is published under the MIT-License. See 'license' for more information.