Turkish Named Entity Recognition: A Survey and Comparative Analysis

Anonymous

Abstract: Named entity recognition is a challenging task that has been widely studied in English. Although there are some efforts for named entity recognition in Turkish language, the reported results are limited to particular datasets and models. Moreover, there is a lack of comparative analysis for named entity recognition in Turkish. In this study, we contribute to the literature in three folds. First, we provide an up-to-date short survey on Turkish named entity recognition studies. Second, we compare state-of-the-art named entity recognition models on various Turkish datasets that we can access to. Lastly, we analyze a set of linguistic processing steps that would affect the performance of Turkish named entity recognition.