Project Description:

Large, easy-to-use semantic networks of symbolic linguistic knowledge such as WordNet and MindNet have become staple resources for semantic analysis tasks from query expansion to word-sense disambiguation. However, the knowledge captured by these resources is limited to formal taxonomic relations between or dictionary definitions of lexical items. While such knowledge is sufficient for some NLP tasks, we believe that broader opportunities are afforded by databases containing more diverse kinds of world knowledge, including substantial knowledge about compound concepts like activities (e.g. “washing hair”), accompanied by a richer set of temporal, spatial, functional, and social relations between concepts.

Based on this premise, we built OMCSNet, a freely available, large semantic network of commonsense knowledge. Built from the Open Mind Common Sense corpus, which acquires world knowledge from a web-based community of instructors, OMCSNet is presently a semantic network of 280,000 items of commonsense knowledge, and a set of tools for making inferences using this knowledge. In this paper, we describe OMCSNet, evaluate it in the context of other semantic knowledge bases, and review how OMCSNet has been used to enable and improve various NLP tasks. Try OMCSNet NOW via this webform! (Thanks to Stuart Horner and Elliot Turner)

Project White Paper [.doc] [.pdf]

Project Team:

Hugo Liu (hugo@media.mit.edu)
Push Singh (push@media.mit.edu)

Download:

OMCSNet v1.3.1 Knowledge Browser and Commonsense Inference API for Python/Java (For other versions, see below)

[view GPL license]
(commercial licensing, please contact us for more information)

download now for free!

• omcsnet v1.3.1 java api and knowledge browser (6.4mb) [.zip]
• omcsnet v1.3.1 python api and knowledge browser (4.9mb) [.zip]
• omcsnet v1.3.1 text version (2.2mb) [.zip]

Version History

• Bugfixed version 1.3.1 (25 Nov 2003)
  • datafiles can now sit in the current working directory (".") or in the path of either of the "PATH" environment variable
  • fixed bug which prevented correct functioning of two concurrent instances of OMCSNet
  • compatibility with MontyLingua resolved (must use MontyLingua version 1.3.1 or higher)
• New in Version 1.2 (11 Aug 2003)
  • lisp-style predicates.txt input added
  • java API released
• New in Version 1.0 (9 May 2003)
  • MontyTagger added to OMCSNet Toolkit
  • first public release
• New in Version 0.2 (10 Aug 2002)
  • OMCSNet high-level python API released
  • version developed for MAS.964 class final projects
• New in Version 0.1 (29 Mar 2002)
  • Commonsense Robust Inference System (CRIS), a component of CS ARIA by hugo@media, is renamed to OMCSNet

Platforms and Compatibility

• The v1.3.1 Java knowledge browser is believed to only run on the Windows platform because it requires a precompiled windows .DLL for the Tkinter GUI package. Edison Thomaz has written a browser for MacOS X (sell below)
• The v1.3.1 Java api has no known operating system dependencies
• The v1.3.1 Java package has been tested with java versions java2 1.3.1 through java2 1.4.1 on the windows and linux platforms.
• The v1.3.1 python api and knowledge browsers have been tested with Python versions 2.2 and 2.3, and have been tested on the windows and linux platforms
• Please send bug reports to us.

Additional OMCSNet Developer Tools

Members of the OMCSNet Development Community have developed some fantastic tools which support and extend OMCSNet. We've linked to these tools for your convenience.

OMCSNetCPP - OMCSNetCPP is a C++ API and inference toolkit for accessing OMCSNet. Written by Elliot Turner.

NEW! OMCSNet Browser for MacOS X - A simple Cocoa-based knowledge browser for OMCSNet. The package includes source code, project files, and OMCSNet data files. It shows how to incorporate OMCSNet within an Objective-C app. Written by Edison Thomaz.

Web / CGI Interface to OMCSNet - Query the OMCSNetCPP Semantic Network and inference tools from a web form / cgi-bin! Written by Stuart Horner, using OMCSNetCPP.

NEW VERSION! OMCSNetCPP+WNLG - A wordsense disambiguated version of OMCSNet, merged with WordNet data, and annotated with part-of-speech information. Written by Elliot Turner

Development Community:

OMCSNet is the quickest way to develop commonsense-informed applications. The following projects have resulted from OMCSNet. We really value your feedback. If you have used OMCSNet in an interesting way and would like to include your testimonial here, please email us.

Projects Using OMCSNet

• Common Sense Conversations: Understanding Casual Conversation using a Common Sense Database (Nathan Eagle, Push Singh, Alex Pentland, Human Design Research Group, MIT Media Lab)

• Smart Conversation Turn-taking in SAM (Austin Wang, Gesture and Narrative Language Group, MIT Media Lab)

• Globuddy, a Dynamic Broad Context Phrase Book (Rami Musa, Madleina Scheidegger, Andrea Kulas, Yoan Anguilet, MIT)

• email us to add your project here