On Detecting Noun-Adjective Agreement Errors in Bulgarian Language Using GATE

Abstract: In this article, we describe an approach for automatic detection of noun-adjective agreement errors in Bulgarian texts by explaining the necessary steps required to develop a simple Java-based language processing application. For this purpose, we use the GATE language processing framework, which is capable of analyzing texts in Bulgarian language and can be embedded in software applications, accessed through a set of Java APIs. In our example application we also demonstrate how to use the functionality of GATE to perform regular expressions over annotations for detecting agreement errors in simple noun phrases formed by two words - attributive adjective and a noun, where the attributive adjective precedes the noun. The provided code samples can also be used as a starting point for implementing natural language processing functionalities in software applications related to language processing tasks like detection, annotation and retrieval of word groups meeting a specific set of criteria.