Related Projects


Xuhua Chen has written An English Dictionary with AB Index and Frequency which uses PAPC (per article per count) basis to count word frequency and two digit numbers to show the frequency rank.

Related Projects

Semantic networks

  • BabelNet is a very large multilingual lexical database, developed by Roberto Navigli  [email] , Simone Paolo Ponzetto and others at the University of Rome "La Sapienza" in the context of the ERC Starting Grant MultiJEDI. BabelNet provides a wide-coverage knowledge repository in which WordNet is automatically aligned to the English Wikipedia, and lexicalizations for its concepts (i.e. English Wikipedia pages and WordNet synsets) are provided on the basis of Wikipedia cross-language links and the output of a machine translations system. The current version, which is freely available under a Creative Commons license, comprises a network of more than 3 million synsets and 70 millions semantic relations, and it can also be browsed through an on-line interface.
  • The Global WordNet Association is a free, public and non-commercial organization that provides a platform for discussing, sharing and connecting wordnets for all languages in the world.
  • MultiWordNet, developed by Luisa Bentivogli [email] and others is a multilingual lexical database, developed at ITC-irst, in which the Italian WordNet is strictly aligned with Princeton WordNet 1.6. The current version includes around 44,400 Italian lemmas organized into 35,400 synsets which are aligned, whenever possible, with their corresponding English Princeton synsets. The MultiWordNet database can be freely browsed through its on-line interface, and is distributed both for research and commercial use. Information on the distribution licence is available at the web site.
  • The Open Multilingual WordNet is a massively multilingual database that links many different WordNet projects, developed by Francis Bond [email] at Nanyang Technological University.  It currently links open-source wordnets in over 20 languages, and allows you to both look words up online or download the data.

Web Interfaces

  • Roger Brideaux has developed Snappy Words, a visual online dictionary based on WordNet data.
  • Duncan Gawn has developed lengusa, a machine learning-powered sentence search engine online. Its cluster of background services constantly collects, sanitizes, indexes, and orders sentences with state-of-the-art algorithms and machine learning routines.
  • Steven Vercruysse, from NTNU University in Trondheim, Norway, has developed an advanced web-interface to browse the WordNet database. It looks and feels like the well-known Visual Thesaurus interface, but it is built on the latest web-technology (pure JavaScript, HTML5). It allows more user-interaction, and it is free to use.
  • A group of students in Hamburg, Germany has created an english-german dictionary, featuring a intuitive interface with synonym, encyclopedia and related words search functions. It is written in PHP and uses a mySQL Database
  • provides antonyms, translation to many languages, voice pronunciation, nearby terms and community related features that let users discuss language related issues and ask questions, as well as international versions of WordNet including Japanese, German and Arabic.
  • Bee Dictionary has created a Web 2.0 English Dictionary with audio, phonetics, usages and other informations using WordNet as base for definitions. Users can save the words in their personal wordlist and practice them as flash cards anytime.
  • Mexidex is a fast, simple, online dictionary and thesaurus based on WordNet 3.0. Added features include complete inflections, auto-suggest, adult-filtering, a browsable index, frequent updates, and support for small-screen, handheld/mobile devices. Developed by Serge Bohdjalian at Memidex [email].
  • Jack Dermody [email] has developed a new web interface that is quite fast, and has the following advantages – query expansion with automatic spelling corrections and the inclusion of the Princeton WordNet Gloss Corpus in the definitions.
  • Another simple web interface for WordNet 2.0 was developed by Nikolai Golovchenko. It returns all of the search results on the same page, with hyperlinks on each word for easy browsing.
  • Philippe Martin [email] converted WordNet 1.7 nouns into an ontology browsable and exploitable for knowledge representation and knowledge sharing by the knowledge server WebKB-2.
  • An interactive CGI interface from which you can select many searches at once was developed at E.N.S.T in France by Francois Yvon, Didier Verna and several undergraduate students. This interface is in French.
  • A hyper-dictionary, HyperDic Online, was developed by Eric Kafe [email].
  • Martin Laplante [mailto] has developed LookWAYup, an online dictionary based on WordNet and other soures. It translates between English and French, Spanish, German, Dutch, or Portuguese. It uses morphology and corrects misspellings. It presents related terms as well as cousins and generates usage sentences from frames. Can be used online or plugs into Netscape, IE, Opera, Palm, various PDAs and mobile phones.
  • Chris Cargile has developed three separate interfaces for retrieving WordNet data in HTML, XML and JSON (with links to example results shown). Further details on the interfaces and the structure of the results available through them is provided here.
  • includes definitions from 4 dictionaries (WordNet, Webster's Revised Unabridged Dictionary, Century Dictionary and Cyclopedia and Chambers's Twentieth Century Dictionary), illustrations, example sentences, interesting facts, quotations, idioms and etymology as well as related words (synonyms, hypernyms, hyponyms).


  • Matt Gerber of Michgan State University has written a C# API for WordNet 3.0. You can read about it here. The API's most distinguishing feature is the option to use either in-memory or disk-based retrieval of the WordNet data. The former is essential for high-performance applications where plenty of memory is available (~200MB). The latter uses essentially no memory and is great for small-footprint applications (it is slower, of course).
  • François-Régis Chaumartin [email] (frc at has developed at Proxem( the Antelope .NET linguistic framework. Antelope uses and extends WordNet; it also includes a subset of a French WordNet and several taggers and parsers. It also implements many out-of-the-box algorithms: similarity measures, Word Sense Disambiguation, Semantic Role Labeling and anaphora resolution. Antelope works under Windows (.NET) or Linux (MONO). It is free for research and education.
  • Troy Simpson [email] has developed a C# interface to WordNet based on Malcolm Crowe's original port, but using WordNet 3.0. It requires the .NET framework and can be downloaded here.
  • Derik J. Palacino [email] has developed a C# interface. This project was built on the database files for WordNet 3.0 however the code was translated from version 2.1. It can be downloaded as a Windows binary or as source code.


  • WordNet TreeWalk, an interface to WordNet based on tree-views developed by Bernard Bou [email], Lycée Champollion, France, has been updated to support WordNet 3.0. A COM component object that can also be accessed from COM-enabled languages (scripting languages, and C++).


  • extJWNL (Extended Java WordNet Library), developed by Aliaksandr Autayeu, is a Java API for creating, reading and updating dictionaries in WordNet format. extJWNL is an upgraded version of JWNL. Open source, Java, available at

  • University of Rochester Computer Science Department has released its URCS WordNet Browser, a cross-platform browser for the WordNet 3.0 database. Implemented in Java, the browser runs on Windows, Mac OSX, and Linux (or any other platform that supports Java), and includes platform-specific customizations that make the user experience as natural as possible. The non-browser classes may have pedagogical value for others interested in using WordNet from Java.

  • WNEJB, by Bernard Bou [email], provides scalable, fault-tolerant, distributed access to WordNet in a Java Enterprise context. A session EJB is the favored interface for queries that return XML strings or serialized W3C documents. WNEJB comes with a deployment wizard, XML servlet and GUI browser, and must be hosted on a JBoss 5 or 6 application server.
  • Bernard Bou [email] has written WNJN, a JNI-compliant Java Native library that interfaces the WordNet database (1.7.1 and 2.0) to Java programs. It offers Java classes that interface with binary code (written in C++) and provides query and search funcationality.
  • A new version of JWNL (Java WordNet Library), written by John Didion [email], has been released. It is a Java API for accessing WordNet, and provides API-level access to WordNet data. It is pure Java (uses no native code), so it is completely portable, and all of the source code is available.
  • JWI (the MIT Java Interface to Wordnet) is a small, fast, full-featured, easily-extensible Java library that supports access to Wordnet 1.6 through 3.0 as well as other Wordnet variants. The library allows mounting dictionary files directly on-disk for a small memory footprint, or loading all of Wordnet into memory for lightning-quick access. The library is compatible with Java 1.5 and up, and is distributed under a non-commercial license with full source code and a short users' guide. JWI was developed by Mark Finlayson.


  • Ramu Ramamurthy has developed an OCaml API to WordNet that allows programs in the OCaml language to use WordNet forms and definitions.



  • A Perl extension module for accessing and manipulating WordNet has been developed by Dan Brian [email]. This module allows access to the Wordnet lexicon from Perl applications, as well as manipulation and extension of the lexicon.
  • Dan Brian [email] has also developed a lightweight Perl module that does a fast binary serach of the indexes. It takes a sense key (as used in the sense index) or a keyword and returns synset information from the corresponding data file.
  • Another Perl interface has been developed by Jason Rennie at the MIT AI Lab. It is object-oriented, grabs WordNet information directly from the database files and allows access to the entire WordNet lexicon.


  • A simple PHP interface to Wordnet 1.7.1 developed by Mike McDonald [email]. It allows PHP scripts access to all the WordNet API functions.


  • Eric Kafe [email] has released Prolog versions of WordNet 3.0 and 3.1 bundled with the original WNprolog-3.0 documentation.


  • Natural Language Toolkit has taken over the development of pywordnet. There is now a Python package, nltk_lite.wordnet, which incorporates pywordnet and which supports WordNet 2.1. It is included in NLTK Lite.


  • has created free APIs based on REST calls which return a well-formatted XML result, providing both synonyms and definitions APIs based on the WordNet database.


  • WordNetScope, by Bernard Bou [email], is an easy-to-use cross-platform browser application for the WordNet SQL database (MySql, PostgreSql, SQlite,...) that displays results as trees. Standard query capabilities extend beyond words to other WordNet entities such as senses, synsets, semlinks, lexlinks, etc. Behind the scenes, the Hibernate framework maps relational data to objects.
  • WNSQL, by Bernard Bou [email], provides a ready-to-use WordNet SQL database in MySQL, PostgresSQL, Sqlite, HsqlDB (cached and memory tables) and Derby.
  • WNPOJO, by Bernard Bou [email], is a library of Plain Old Java Objects to encapsulate WordNet entitites (word, sense, synset, etc.) and handle them transparently without having to manage SQL queries, using the Hibernate framework. Comes with a programming sample.
  • A companion project to WNPOJO, WNTRANS comes with a GUI wizard and currently supports MySQL, PostgresSQL, Sqlite, HsqDB, and Derby, but this set is extensible. WNTRANS provides portability to the WordNet SQL database through the Hibernate framework.
  • [MySQL, PostgreSQL] A ready-to-use SQL database that unifies WordNet 3.0, WordNet 2.0-2.1, 2.1-3.0, 2.0-3.0 sensemaps, VerbNet 2.3, XWordNet 1.1 compiled by Bernard Bou [email] supports both MySQL and PostgreSQL.
  • [MySQL, PostgreSQL] WNSQLBUILDER is a Java tool by Bernard Bou [email] to build a unified SQL database (see above) from the WordNet, WN sensemaps, VerbNet, XWordnet,, BNC and SUMO standard releases.
  • [MySQL] Android Technologies, Inc. provides a MySQL version of WordNet 2.0 converted from the Prolog files.
  • [MySQL, dBase] wordnet2sql, developed by Richard Bergmair [email] is a system that converts WordNet 3.0 into SQL.


  • Troy Simpson has written WordNetDT, an alternate to WNB with some advanced search features.


  • Bernard Bou [email] has also written a Java servlet to be hosted in a Tomcat server conainer which delivers XML output to queries. A DTD is provided to format XML query results.


  • IDSIA, developed by Simone Franza [email], that represents the searched term with the corresponding hypernyms, hyponyms, etc.
  • WordNet::Similarity, developed by Ted Pedersen [email], is an open source Perl module for measuring the semantic distance between words. It provides a number of measures of semantic similarity and semantic relatedness based on WordNet. Given two synsets, it will return a numeric score showing their degree of similarity/relatedness according to various measures that all rely on WordNet in different ways. It provides support for estimating information content values from untagged corpora, including plain text, the Penn Treebank, or the BNC. 
  • WordQuake is a Chrome extension that transforms webpages using WordNet as a source of creative new words. From the author, Alan Du, "Tired of what you see on a webpage? Want to spice things up? Or surprise your friends? Introducing WordQuake, an application that transforms entire webpages before your very eyes. Create new meanings from online news articles, editorials, classic literature...even your best friend's Facebook posts." Alan developed WordQuake with two other Princeton undergraduate students, Keith Gladstone ([email protected]) and Derrick Xin ([email protected]).
  • WordNet::Similarity requires a local copy of WordNet as well as the WordNet::QueryData Perl module, which provides a direct Perl interface to the WordNet database. Author Jason Rennie also has other WordNet tools on his site.
  • "A Complete Multilingual WordNet List by Language" was compiled by Samuel Chong of Pasadena City College.
  • Indio, by Dirk Holland, is a Microsoft Word Add In based on Wordnet 3.0. Simply highlight a word in a document to get its meaning. Words can also be typed in via the form. Indio also includes wild card searches and other advanced features such as the ability to click on lexical or semantic relations to get their definitions. Indio also includes definitions from Webster's 1913 dictionary. Availalbe via aemail directly from the author.
  • Joshua Harwood has produced a spreadsheet from WordNet which makes a frequency list for some entries (selected from other frequency corpora), and the tokens that are contained within the most frequently occurring entries. 
  • How do you pronounce English words and use them? Check out the EmbedPlus YouTube embed tool, which uses WordNet definitions and word metadata for its video pronunciation API for dictionaries. Start here by watching videos of how to pronounce gyro, salmon, and meme; three cases of often mispronounced words. Contact them for more info.
  • Adam Pease has announced that WordNet is now available in OWL with links to the Suggested Upper Merged Ontology as a single large OWL file and dynamically served as OWL pages comprising single synset at a time (e.g.physical_object)
  • Dinesh Babu has developed a WordPress plugin, Click2Refer, which lets blog readers look terms up in Wordnet just by a double clicking on the word.
  • Artha is a cross-platform open thesaurus with WordNet as its database for Linux/Unix-like systems. Once the application is launched it sits on the system tray monitoring for a set hot key combination. The user can select text on any window, and call Artha by pressing the set hot key combination. It can also show notification of the selected word's prime definition. Artha is developed and maintained by Sundaram Ramaswamy [email].
  • JWord is a WordNet-inspired tool that integrates other word sources such as Roget's Thesaurus. It includes a simple browser and, for developers, the ability to integrate other sources.
  • Panagiotis Toulis [email] has developed a cool little Perl interface that does a WordNet lookup.
  • DefineItFast is an AJAX & PHP Driven online dictionary that is powered by Wordnet 3.0. Definitions are shown interactively as the user type the word.DefineItFast was developed by Paolo Medina [email].
  • Armando Stellato has written a plugin called OntoLing for the Ontology editing tool Protégé It allows the user to: browse linguistic resources, enrich ontologies with elements from these resources, and build new ontologies starting from existing resources. The WordNetInterface for this tool is based on the JWNL library, and is compatible with WordNet 1.6 through 2.0.
  • Bernard Bou [email] has written WNWA, a WordNet Web Application to be hosted on a Java Enterprise Bean Container (typically JBoss). It includes an Enterprise Java Bean that interfaces with the WordNet SQL database. Tomcat front-end servlets offer 4 renderings of query results: xml document, HTML, DHTML tree, hyperbolic tree. A demo site with limited hardware/bandwidth capabilities has been deployed, but its continued existence is not guaranteed.
  • Tony Veale of University College Dublin is heading KNOW-BEST, which is attempting to use WordNet and other lexical ontologies as the generative basis for on-line language games. Playable samples of the games and information/papers related to their construction/research issues are available at The Creative Language Systems Group web site.
  • Keith Alcock has developed the WordNet Relationship Browser, a Windows program that allows a user to browse related words through the various links in WordNet. Each path thus created is given a score based on user-configurable weights. It is based on a version of WordNet contained in a .dll file, and is thus extremely fast.
  • Sandiway Fong [email] has released wnconnect, Prolog-based software which given two terms, does a breadth-first search for the possible paths between them in WordNet 1.7.1. There is a MacOS X GUI application with Sicstus Prolog embedded in it. It generates graphs of the connections. There is also a TTY version with just the Prolog and C files. You do need Prolog for this version.
  • The Suggested Upper Merged Ontology (SUMO) is an ontology specified in first order logic. SUMO mappings to WordNet 3.0 are available. Email Adam Pease for more information.
  • A Resource Description Framework (RDF) representation of WordNet and ontology defining the terms used to represent the RDF version were developed by Sergey Melnik and Stefan Decker [email].
  • Steve Reed [email] has converted WordNet 2.1 to RDF (Resource Description Framework). It is available from his web site Texai:
  • Ken Litkowski [email] of CL Research provides an alphabetic verion of WordNet 2.0. There are 143991 entries in this dictionary, with a sense for each occurrence of an entry in a distinct synset. Virtually all information in WordNet has been captured, including the new domain relations, verb groups, and derivational forms. 
  • Treebolic offers a hyperbolic representation of a hierarchy of data. A browser is provided which offers a hyperbolic view of WordNet words and links. The user can browse through WordNet trees/nets by dragging the word or data nodes.
  • A semantic rhyming dictionary was developed by Doug Beeferman at Carnegie Mellon University. It uses WordNet to help sort the output based on how near in meaning a word is to a certain target meaning. He has recently added synonym and semantic sibling queries to the interface.
  • Doug has also developed Lexical Freenet, a Web-based thesaurus and word discovery/connection program.
  • WordWeb 5,based on WordNet 3.0, is a free thesaurus/dictionary for Windows based on WordNet, developed by Antony Lewis.
  • Automatic Semantic Annotation of Texts, available via the home page of Fernando Gomez, is a small semantic corpus of 500 sentences, in which semantic roles, verb predicates and noun senses are resolved, using WordNet 1.6. The 500 sentences contain about 400 verbs.



  • WordNetPort, developed by Josephine E. Petralba [email] provides distribution copies of WordNet for loading into databases. A collection of SQL scripts and a batch file is available for easy loading into an Oracle database. Files in Excel and other formats are also available for Non-Oracle users. Demo queries are found at the home page. Most objects are created with reference to the WordNet dictionary displays, where the aim is to come-up with simple queries whose results are similar to the content displayed online.
  • The app Kamusi Here!, developed by Martin Benjamin [email], is a translating dictionary that incorporates WordNet, and works in more than 11 languages.