Automatization of Valency Frames Prediction Based on Verb Prefixation with NooJ

master's thesis

• Author: Petra Bajac
• Mentor: Božo Bekavac (mentor)
• Mentor: Krešimir Šojat (komentor)
• Committee member: Kristina Kocijan (predsjednik povjerenstva)
• Granter: University of Zagreb; Faculty of Humanities and Social Sciences; (Department of Linguistics); Zagreb
• Defense date and country: 2022-03-23, Croatia
• Scientific / art field, discipline and subdiscipline: HUMANISTIC SCIENCES; Philology; General Linguistics

Abstract

By approaching the phenomenon of verbal prefixation in Croatian through the verb dati ‘to give’ and its 18 derivationally related forms, such as prodati ‘to sell’ and dodati ‘to add’, this thesis outlines the impact prefixation can have on the verb’s valency frames. By utilizing the NooJ application to create a formal grammar which can detect and tag derivationally related verb forms, and with the help of a model <NP>/<PP> chunker for Croatian, our research aims to start the process of automatization of valency frames prediction and thus to speed up the analysis of prefixed verbs and their valency frames. Our analysis is focused specifically on those verbs which are not currently present in the Croatian valency lexicon CROVALLEX, and as a result, we provide the valency frame descriptions for 10 missing verbs from the dati derivational family. Furthermore, this thesis provides a template for the automatized creation of connections between derived and base verb forms.

Abstract (croatian)

Pristupivši procesu glagolske prefiksacije u hrvatskom jeziku bazirajući se na primjeru glagola dati i njegovih 18 derivacijski povezanih oblika, poput prodati i dodati, ovaj rad prikazuje utjecaj prefiksacije na promjene u valencijskim okvirima derivacijski povezanih glagola. Cilj je ovog rada započeti proces automatizacije izrade valencijskih okvira pa smo tako izradili model formalne gramatike putem aplikacije NooJ koja može pronaći i označiti derivacijski povezane glagole. Uz pomoć modela razdjelnika za imenske fraze i prijedložno-padežne izraze, naš pristup ubrzava proces analize prefigiranih glagola i pripadajućih okvira. Naša se analiza primarno fokusira na one glagole koji trenutno nisu prisutni u CROVALLEX-u, valencijskom leksikonu glagola hrvatskoga jezika, a koji su dio derivacijske mreže glagola dati, te smo stoga opisali valencijske okvire 10 prefigiranih glagola koji nedostaju. Također, kao rezultat ovog rada nastao je i mogući predložak za automatsku izradu poveznica među deriviranim i osnovnim oblicima glagola.

• Keywords: prefix; prefixation; derivation; valency frames; computational lexicons; NooJ
• Keywords (croatian): prefiks; prefiksacija; derivacija; valencijski okviri; računalni leksikoni; NooJ
• Language: english
• URN:NBN: urn:nbn:hr:131:708196
• Study programme: Title: Linguistics (double major); specializations in: General Linguistics, Comparative Linguistics, Computational Linguistics, Aplied Linguistics and Cognitive Linguistics, Cognitive; Course: Computational Linguistics; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra lingvistike (magistar/magistra lingvistike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2022-03-30 16:33:20