Research Area:  Machine Learning
Named entity recognition (NER) has been considered as an initial step for many applications and tasks such as information retrieval and extraction, question answering, topic modelling, open information extraction, knowledge graph construction, and so forth. Therefore, NER has been receiving increasing attention in the research community. Despite the abundant availability of previous studies on NER, few of them have been applied for more than one dataset. Hence, one system might outperform other systems in one dataset and fail to do in another one. The previous NER surveys have mostly focused on reporting the NER systems without providing a clear comparison for all systems proposed for each dataset. In this paper, we will track the NER performance progress for the most commonly used datasets in NER and report the most recent best systems that have been proposed for each dataset during the last few years.
Keywords:  
Author(s) Name:   Mohammed Albared; Marc Gallofré Ocaña; Abdullah Ghareb; Tareq Al-Moslmi
Journal name:  
Conferrence name:  First International Conference of Intelligent Computing and Engineering (ICOICE)
Publisher name:  IEEE
DOI:  10.1109/ICOICE48418.2019.9035170
Volume Information:  
Paper Link:   https://ieeexplore.ieee.org/abstract/document/9035170