Advanced Developments in Arabic Named Entity Recognition: A comprehensive Study

Abstract

Why has extracting important and essential information from Arabic texts become so useful and necessary? The answer to this question lies in the fact that the frequent appearance of Arabic words and texts on the Internet has led to interest in this topic. Named entity recognition (NER) refers to a fundamental task that is an integral part of many natural language processing (NLP) functions, such as information retrieval and machine translation, which are tasks of information extraction. When we review previous research and studies, we see that they relied on the recognition of known entities (NER) from large sources of knowledge and the application of hand-made features. This approach takes a long time and is no longer sufficient for languages with scarce resources, such as Arabic. Recently, the process of recognizing named entities in the Arabic language (NER) has begun to attract attention. The features and characteristics of the Arabic language, as a member of the Semitic language family, have become major challenges to recognizing named entities. The performance of the Arabic NER component positively impacts the overall performance of the NLP system. Many researchers have improved the methods to extract a variety of entities from different text types and languages. Additionally, there has been a push in the research community to update, develop, and implement new strategies that are considered more modern and innovative for extracting diverse and different entities that contain useful names in various natural language applications. In this paper, we provide an overview of the research advancements and progress made in classification studies and Arabic Named Entity Recognition.