Efficient Named Entity Recognition on Greek Legislation

Abstract

The legal landscape is constantly becoming more and more complex. Artificial Intelligence and Machine Learning are starting to shine in many fields, so it is only natural that we examine their implementation in the Legal Domain as well. While many researchers are investigating the application of these techniques in English, for low-resource languages like Greek, the field is far from explored. Our work is a step in that direction. We describe our Named Entity Recognition solution to Greek Legislation and compare it with the available state-of-the-art Greek implementations and present how our approach outperforms them.