jnulogo

HOME

Introduction Dissertation Sample Verb Phrase GA Phonemic Inventory Entering Text System Limitations Future Work FAQ

Back to the Analyzer
 
 

The Great Andamanese Verb Analyzer

Welcome to the Andamans and its indigenous people...

We will help you go to the sediments of the human history and tell you how and what they know. Through their language...

You are on a page which will lead you to the Great Andamanese Verb Analyzer, a verb analyzer for one of the least spoken languages of the world.

Besides it also contains other kinds of information on the people of Great Andamanese, the work done and going on in this area, as well as many links on this topic.

But before this, let me introduce you to the main thing why I am here. The whole idea of appearing on the web comes from the dissertation I wrote. It was titled “Developing a Computational Framework for the Verb Morphology of Great Andamanese”.

The name of the technology developed has been given as Great Andamanese Verb Analyzer or GAVA for short. It analyzes the verb phrase from a given text in Great Andamanese. The nature of the technology developed is that of a knowledge system which can also function as an aid to machine translation and other NLP applications. In other words, it can be at the levels of a product, a technology and as a sub-system. The program is an online Java/JSP application running on Tomcat-Apache webserver. It has a JSP front end and text files as back-end. It can be hosted on any platform and can be viewed as HTML. The input and output is in IPA UTF-8 (refer to the Appendix I).

This program has been developed as a web application using Java/JSP on Apache Tomcat platform. As the language does not have any script, IPA characters provided by Lucida Sans Unicode font have been used. Any Unicode font supporting IPA can be used to view and input the texts. However the technology is fully portable from one system to another as only text files for database has been used and the tools used are platform independent.

The program is highly scalable and can be easily extended to work as text processor and tagger for Great Andamanese or languages with structural similarities.  Most of the documentation for the program part has been completed. Testing of the software is currently undergoing and we are getting more than 90% results. An appendix (IV) to this effect has also been attached showing the result of the most frequent verb phrases in the basic sentence list.