Unit:
Department of Turkish Studies and Modern Asian StudiesLibrary of the Faculties of Political Science and Public Administration, Communication and Mass Media Studies, Turkish and Modern Asian Studies, Sociology
Dissertation committee:
Ιωάννης Ε. Σαριδάκης, αναπληρωτής καθηγητής, Τμήμα Τουρκικών Σπουδών και Σύγχρονων Ασιατικών Σπουδών, ΕΚΠΑ.
Αγγελική Φωτοπούλου, Διευθύντρια Ερευνών, Ινστιτούτο Επεξεργασίας του Λόγου, ΕΚ ΑΘΗΝΑ
Γεώργιος Μικρός, καθηγητής, Τμήμα Ιταλικής Γλώσσας και Λογοτεχνίας, Εθνικό και Καποδιστριακό Πανεπιστήμιο Αθηνών
Ζωή Γαβριηλίδου, καθηγήτρια, Τμήμα Φιλολογίας, ΔΠΘ.
Παναγιώτα Κυριακοπούλου, καθηγήτρια, Institut Gaspard Monge, Université Paris-Est Marne-la-Vallée
Γεώργιος Μαρκόπουλος, αναπληρωτής καθηγητής, Τμήμα Φιλολογίας, ΕΚΠΑ.
Ανδρομάχη-Βιργινία Πανταζάρα, επίκουρη καθηγήτρια, Τμήμα Γαλλικής Γλώσσας και Φιλολογίας, ΕΚΠΑ.
Original Title:
Το σημασιολογικό πεδίο των συναισθημάτων: ταξινόμηση των ρημάτων της Νέας Ελληνικής που δηλώνουν συναίσθημα
Translated title:
The semantic field of emotions: A Lexicon-Grammar account of Greek verbs denoting emotion
Summary:
This doctoral dissertation presents research aimed at the study of Greek verbal predicates denoting emotion, and their classification into Lexicon-Grammar tables. The ultimate purpose is the creation of a robust lexical resource that is suitable for translation – performed by humans and machines alike. The linguistic analysis follows the theoretical assumptions of transformational and distributional grammar introduced by Zelling S. Harris (1951, 1954, 1968, 1981), whereas the representation of the verbs abides by the principles set by the Lexicon-Grammar framework proposed by Maurice Gross (1975). Being a model of syntax limited to the elementary sentences of the form Subject – Verb – Object, the theory argues that the unit of meaning is located at the sentence rather than the word level. In this respect, our study is based on empirical data drawn from material collected from electronic corpora (corpus-based).
In order to better delineate the linguistic data, the definition of the semantic field of EMOTION was in order and the verbs denoting emotion were collected. Subsequently, a corpus of text documents was compiled in view of eliciting real-world data on the syntactic and semantic properties of the verbal predicates at hand. Ultimately, the verbs studied were classified into five (5) Lexicon-Grammar tables, based on a shared definitional structure and shared syntactic and semantic-distributional properties. At the next stage, in view of better representing the emotion concept each verb denotes, additional semantic properties were defined, which correspond to a set of pre-defined attribute/value pairs. We argue that these pairs define a taxonomy (annotation scheme) that better depicts the semantic class of EMOTION. The values assigned to the verbs at hand were encoded as properties in supplementary Lexicon-Grammar tables.
At the final stage, a cascade of graphs that comprise a computational grammar was manually crafted using a lexicon- and grammar-based corpus processing suite. The grammar makes use of the information encoded in the Lexicon-Grammar tables, rendering itself appropriate for validation purposes. The final grammar can be integrated into a future syntactic parser or a system for emotion detection.
Main subject category:
Social, Political and Economic sciences
Other subject categories:
Language – Literature
Keywords:
Verbal predicate denoting emotion, psych verb, Lexicon-Grammar, distributional properties, syntactic properties, semantic properties, transformation, graph, local grammar, Natural Language Processing, emotion detection.
Number of references:
176
