Skip to content

WordTies presents a search interface and visualisation of the conceptual network of a WordNet, to browse its relational and multilingual links to other wordnets. Originally developed for visualising DanNet, the Danish wordnet.

License

Notifications You must be signed in to change notification settings

kuhumcst/wordties

 
 

Repository files navigation

WordTies

WordTies, is a browser application to view the conceptual network of a Wordnet and to browse its relational and multilingual links. This is a fork of Anders Johannsen's AndreOrd. AndreOrd project has been renamed and deployed as WordTies for the METANORD project.

See the description on the CLARIN ERIC Portal - Showcases > Wordties page.

New Features

  • D3.js visualation/graphs (migrated from Protovis)
  • Multilingual linking
  • Improved Alignment importing via Python script
  • Alignment search by CorePWN or Multilingual alignments
  • Danish and English (locale) translation
  • Build/import scripts

Requirements

PostgreSQL

See PostgreSQL Installation Guides

Rails (3.2.x)

See Installing Rails

Python

Node.js

See Node.js Wiki - Installation

Installation

It is recommended to install on a standard Linux distribution.

Setup PostgreSQL database

  • Create user createuser -U postgres -P -S andreord
  • Create database createdb -U postgres -O andreord andreord_db
  • Import schema psql -U andreord -d andreod_db < db/development.sql

Database & locale configuration

  • Edit 'config/database.yml' and set database and username
  • Edit 'config/application.yml' and set desired locale config.i18n.default_locale = :en

Import Wordnet data

Via Rails console in your working directory ($RAILS_APP) $ rails console:

d = Import::DanNetImporter

d.import

Import Alignments (Optional)

For cmd-line usage help type: python lib/import/princeton_links.py --help

  • Import English CorePWN links (see example) python lib/import/princeton_links.py -f pwn_data/DanNet/eq_core.tsv -n 1000 -v
  • Import multilingual alignments (see example) python lib/import/princeton_links.py -f pwn_data/FinWN/FinnDannValidation_final_import_ext_syn_2.tsv -s fiwn20 -t wordnet30 -l fi -u /wordties-fiwn -n 1000 -m 1 -v

Deployment

Standalone

  • WEBrick Web Server
    • In your working directory ($RAILS_APP), running the app locally on your machine with rails server will start the app at: localhost:3000

Apache /w Phusion Passenger

Demonstration

See the current live demonstration on wordties.cst.dk.

About

WordTies presents a search interface and visualisation of the conceptual network of a WordNet, to browse its relational and multilingual links to other wordnets. Originally developed for visualising DanNet, the Danish wordnet.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Languages

  • Ruby 57.1%
  • JavaScript 13.1%
  • HTML 12.6%
  • PLpgSQL 9.0%
  • CoffeeScript 3.0%
  • Python 3.0%
  • CSS 2.2%