/Smarty

The most advanced context-aware NLP-powered word-sense disuambiguation dictionary in mid-late 2000s before the powerful machine translation - still useful. Extendable Framework for Bilingual and Multilingual Comprehension Assistants in C#

Smarty

Extendable Framework for Bilingual and Multilingual Comprehension Assistants in C# by Todor Arnaudov - Tosh http://twenkid.com

An intelligent dictionary, designed to integrate WordNet, BalkaNet and an ordinary dictionary databases with a rich GUI, tooltips and various NLP processing operations.

Originally developed in C# from scratch, about 9000 lines of code, written for about 3 months in the spring of 2007, during an undergraduate internship at the Wolverhampton University's Research Group in Computational Linguistics.

  • The binary application, the 2008 version 0.1 NB (Not Balkanet) with the original thesis, LREC 2008 and IMCSIT 2008 conference paper, presentation slides:

Version 0.1 NB (Aug 2008) Re-release on 14.2.2021

https://github.com/Twenkid/Smarty/blob/master/Smarty_0-1NB-2008-Comprehension-Assistant-Release-2-2021.zip

Dictionaries load automatically, a bigger text area, fixed bugs from the 2007 version etc. Definitions of the multiword expressions in a tooltip etc.

Known bug: Sometimes search endings doesn't show words which matched exactly the search, e.g. "ed" would not return "ed", but only longer words.

See also the Readme in the archive.

Run Smarty.Exe, Select File -> Load Dictionary See Readme.txt and the /README folder.

A newer version of the program existed, released in 2008. It fixed a few bugs and had functions to measure time for queries (if not mistaken) - needed for the paper. The link was gone after migration, I'll try to dig it out.

http://artificial-mind.blogspot.com/2008/07/smarty-extendable-framework-for.html

  • See also:

https://github.com/Twenkid/Smarty

https://github.com/Twenkid/NLP-Computational-Linguistics

https://github.com/Twenkid/GPT2-Bulgarian-Training-Tips-and-Tools

https://github.com/Twenkid/Suchinitel-1-GPT2-Synthesizer-work

https://github.com/Twenkid/Similarity-NLP-Corpus-CPP