Abstract:

In this study, the aim is to reveal the emotions of the words and compositions in the Turkish languages spoken by Neşet Ertaş.In this direction, the words of 339 works and the knowledge of the composition was obtained. There is a conversation between the words. Three different sources were used to obtain words and speech text. A conversion process has also been implemented for words and non-speech, so text-based or phrase-based texts have been obtained. In this process of conversion, the unnecessary labelling in the texts, such as stops, reflection, question additions, has been transformed into a fragmented, unnecessary. After this phase, phrases or phrases are made processable. Priority for the words is the technique of artificial learning and the technique of miraculous Turkish emotional analysis method is applied and the emotional scores of the words are calculated. Three different artificial learning models have been used to calculate emotional scores. This three-variant model is the state of the models developed by Google adapted to Turkish. Only a sensational value has been obtained in the size of a phrase or phrase. For the purpose of ensuring consistency, this emotional score mediation is calculated from three different models. Then a sensual vector was created consisting of the phrase or phrase of the speech text of each kind. The average of this emotional vector is the average emotional score of that text. The variance of this sensation vector is also calculated. This variance was used to measure the emotional enthusiasm within the style or speech text, and then, for the composition information, the information was based on a music listening platform that was widely used. There are 10 variables in the area of acousticity, dancing, energy, instrumentality, vitality, sound height, speech, pace, value, popularity. Three of these variables have been taken into account, dance-reducable, energy and value variables. The average of these three variables is calculated by the emotional score of the composition. These scores were later matched with the information of the 169 single work data processed, and an atlas of emotions that can be modeled to the word and composition size that shows the emotional map of Neşet Ertaş. Emotions are calculated in the range of axes (-1, +1) in the atlas.Therefore, the emotional scores obtained for both texts and compositions are scaled from the range (-1, +1) by a developed method. Thus the emotional scores of both texts and compositions have been made comparable.For the atlas of emotions after this stage, the texts are placed in the x-excellence. They are placed in the Y-Sense. In the x-exon, called the dictionary, -1 score is indicated with the feelings of "happiness" and +1 score is indicated with the feelings of "happiness". In the isey-exen, the score is -1 "melankolic" and the score +1 is shown with "comfortable" emotions. According to the findings obtained by NeşetErtaş, the dominant characteristics of the texts spoken by NeşetErtaş are found to be based on flawed words, in a large part of the works, the words are flawed and composed in a melancholic structure. It is also found in his words that the emotional changes are characteristically excessive, that they contain emotional enthusiasm; thanks to the word cloud analysis, it is also determined that in his words, in terms of discourse, it is often in human, moral advice.