This diploma thesis is focused on solving the issue of natural language processing, specifically task of named entity recognition in Slovak language. The thesis describes the basic principles and methods used to solve problems of named entity recognition and annotation of text data. Within the practical part of the work, two text datasets are created in the Slovak language. Source for these datasets are articles from the Slovak Wikipedia. An annotation scheme is prepared, and annotations of text data are performed using the Prodigy system in order to further use them for training of the model. The main goal was to train a statistical model for named entity recognition in Slovak language, which is realized by various methods using the spaCy library. The result of a series of training experiments is a model capable to recognize Slovak named entities in four predefined categories with a total score above 72%. The accuracy of different training methods is compared during the evaluation part. Also, the influence of the size of the training set on the overall accuracy of the classifier is monitored and potential possibilities for improving the accuracy of the trained model are defined and analyzed.