Presenter: Arbana Kadriu
Session: 3
Room: MainRoom

Natural Language Processing is one of the most rapid developing fields nowadays. But, unfortunately this is correct only for those languages that have commercial demand. For the other, minor languages this is not the case. So, they must rely on open-source toolkits and built-in software packages. For the most part of languages, pools of raw texts or even simple wordlists are the resources most urgently needed. In this paper, we try to analyze and bring some open source solutions that can help in building NLP applications for a wider spectrum of languages. These solutions include designing text corpora, lexical databases, implementing a complete morphological and sentence parser for a particular language, building language models etc. Open-source toolkits and resources generate frames for research and data maintenance and this makes possible to overcome the isolationism of minor languages. And they play an important role to the advancement of language technology as a research area in general. This relates to the idea of free software that refers to freedom and progress, and not to the price.