Named entity recognition algorithms and applications for non-structural texts

Named entity recognition (NER) problem is considered as a sub-branch of fields such as data extraction, natural language processing and text mining. Named entity recognition is a tool used to assign classes such as predetermined person name, organization name or place name according to the suitability of entity names in unstructured texts. NER studies have uses in many fields. Examples of these are the creation of chatbots, suggesting content on social networks, processing resumes or categorizing customer calls and gaining insights from them, etc. can be said. In this study, NER was performed on two different conditions. Firstly, on a dataset consisting of news articles in English, using two different pre-trained libraries, Spacy and Stanford NLP libraries, the name of the person, the name of the place, the name of the organization, etc. entity names have been tried to be recognized. At the end of this study, the accuracy rates obtained with the libraries, the working structure of the libraries, their speed, etc. criteria were compared. In the rest of the study, using Turkish tweets on Twitter, swearing, insults and inappropriate words were handled as a named entity definition problem and these words were tried to be determined by different methods. First, the words and phrases in the texts were labeled, and then the labeled words were vectorized. Vectors are trained using RNN, bidirectional RNN, GRU, bidirectional GRU, LSTM, bidirectional LSTM and a pre-trained multilingual BERT model. The study results of the models were analyzed and the results of the models were evaluated comparatively.