A study on the use of simulation coefficients for categorical variables in text mining

The transfer of texts into digital environments is facilitated by technological developments. Text mining techniques should be applied in order to reach the texts and to obtain meaningful information from the texts. Text mining aims to obtain statistical results over the text along with the natural language processing discipline from data that is not found in the structure form. Within the scope of this study, by using the text mining techniques and similarity-distance measurements on 6 different works of 4 different authors, the success of author recognition and text classification with the help of R program has been tried to be achieved. The numerical data related to the authors' works were extracted and the Euclidean distance was calculated between these numerical data and the text classification was done by using the K-NN algorithm. It has been tried to determine which text set belongs to a given work, based on the data obtained.