National and Kapodistrian University of Athens
Προσδιορισμός συγγραφικού προφίλ μέσω τεχνικών επεξεργασίας φυσικής γλώσσας
Abstract
dc:descriptionThis thesis presents the research and its outcomes regarding the automatic recognition of author’s personality features based on natural language processing techniques. In particular, we examine the hypothesis that the elements of a person’s character can be determined automatically through the language he/she uses in written speech. For this purpose, an electronic corpus of texts has been developed comprising essays written by high-school students in Modern Greek. Participant students were given two personality questionnaire, one based on the typology of Carl Jung and the other based on the Model of Five Factors (psychological traits). In addition, in this volume, a detailed overview is recorded of international literature of the research field in question, in order to study the methodology, the tools and the results of relevant research up to now. By utilizing modern research findings, the thesis examines the effectiveness of numerous stylometric features to predict the personality of the students. These features, after being automatically extracted from the electronic corpuses of essays using tools and natural languages processing resources, fed a long series of machine learning algorithms, the models of which were later controlled for the accuracy of their efficiency. As regards the Personality Types Questionnaire (Myers-Briggs Type Indicator (ΜΒΤΙ)) we found that the Naive Bayes algorithm renders the highest average accuracy percentage, being 76.5%, while for the prediction of personality features based on the Five Factors model, the Generalized Linear Model algorithm prevailed by average accuracy 72.2%. From the research that we conducted new combinations of stylometric features resulted and corresponding computational techniques, giving interesting and satisfying solutions to the problem of author’s personality automatic recognition for Greek, while the important value of the use of stylometric linguistic features was demonstrated (for Greek as well) in tackling the research problems in the general scientific and technological field of human written speech understanding by computers, emphasizing on the processing of natural language texts for automatic mining of distinct features of their author.
Degree
thesis:*- Grantor dc:publisher
- National and Kapodistrian University of Athens
- Year dc:date
- 2021
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Gagiatsou, Sofia
- Γαγιάτσου, Σοφία
Subjects
dc:subject × 30- Πρόβλεψη προσωπικότητας
- Τυπολογία Jung
- Μοντέλο πέντε παραγόντων
- Υπολογιστική υφομετρία
- Υφομετρικά χαρακτηριστικά
- Μηχανική μάθηση
- Personality prediction
- Jung typology
- Big five model
- Computational stylometry
- Stylometric features
- Machine learning
- Γλώσσα και Λογοτεχνία
- Ανθρωπιστικές Επιστήμες και Τέχνες
- Γλώσσα και Γλωσσολογία
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Φυσικές Επιστήμες
- Διεπιστημονικές εφαρμογές επιστήμης ηλεκτρονικών υπολογιστών
- Εκπαίδευση
- Κοινωνικές Επιστήμες
- Εκπαίδευση και Έρευνα της εκπαίδευσης
- Languages and Literature
- Humanities and the Arts
- Language and Linguistics
- Computer and Information Sciences
- Natural Sciences
- Computer Science Interdisciplinary Applications
- Education
- Social Sciences
- Education and Educational Research
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/49050