Analysis of speech fluency in patients with schizophrenia

master's thesis

Mila Marinković (Author), Jure Žabkar (Mentor), Polona Rus Prelog (Co-mentor)

Abstract

Schizophrenia is a complex mental disorder characterized by significant cognitive and linguistic impairments, particularly in verbal fluency. This thesis addresses the challenge of distinguishing between individuals with schizophrenia and healthy controls by directly testing and examining their verbal fluency performance. To achieve this, we administered established verbal fluency tests to participants, applying these tests in the Slovenian language—a novel approach, as no such analysis had previously been conducted in this language. Our goal was to uncover patterns that could effectively differentiate these two groups within the Slovenian-speaking population. We combined traditional statistical methods, semantic similarity measures, temporal dynamics, and machine learning techniques to analyze word frequency, error patterns, and the organization of verbal output. A key contribution of this research is the successful classification of individuals into schizophrenia and healthy control groups using the Naive Bayes classifier based on verbal fluency features. The study also highlights significant verbal fluency differences between the groups, including higher error rates in individuals with schizophrenia, underscoring the potential of advanced analytical techniques to enhance diagnostic accuracy.

Keywords

verbal fluency;schizophrenia;Slovenian language;machine learning;computer science;master's thesis;

Data

Language:	English
Year of publishing:	2024
Typology:	2.09 - Master's Thesis
Organization:	UL FRI - Faculty of Computer and Information Science
Publisher:	[M. Marinković]
UDC:	004.85:616.895.8(043.2)
COBISS:	210216707
Views:	91
Downloads:	34
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	Slovenian
Secondary title:	Analiza tekočnosti govora pri bolnikih s shizofrenijo
Secondary abstract:	Shizofrenija je kompleksna duševna motnja, za katero so značilne izrazite kognitivne in jezikovne motnje, zlasti v tekočnosti govora. Ta magistrska naloga obravnava izziv razlikovanja med posamezniki s shizofrenijo in zdravimi kontrolnimi osebami z neposrednim testiranjem in preučevanjem njihove tekočnosti govora. Da bi to dosegli, smo udeležencem izvedli uveljavljene teste tekočnosti govora, pri čemer smo te teste uporabili v slovenskem jeziku—kar je nov pristop, saj takšna analiza v tem jeziku doslej še ni bila izvedena. Naš cilj je bil odkriti vzorce, ki bi lahko učinkovito razlikovali med tema dvema skupinama v slovensko govoreči populaciji. Združili smo tradicionalne statistične metode, meritve semantične podobnosti, analizo časovne dinamike in tehnike strojnega učenja za analizo pogostosti besed, vzorcev napak in organizacije verbalnega izražanja. Ključni prispevek te raziskave je uspešna klasifikacija posameznikov v skupine shizofrenije in zdravih kontrolnih oseb s pomočjo Naivnega Bayesovega klasifikatorja na podlagi značilnosti tekočnosti govora. Študija prav tako izpostavlja pomembne razlike v tekočnosti govora med skupinama, vključno z višjimi stopnjami napak pri posameznikih s shizofrenijo, kar poudarja potencial naprednih analitičnih tehnik za izboljšanje diagnostične natančnosti.
Secondary keywords:	tekočnost govora;jezikovne motnje;bolniki s shizofrenijo;slovenščina;testiranja;meritve;statistične metode;magisteriji;Strojno učenje;Shizofrenija;Računalništvo;Univerzitetna in visokošolska dela;
Type (COBISS):	Master's thesis/paper
Study programme:	1000471
Embargo end date (OpenAIRE):	1970-01-01
Thesis comment:	Univ. v Ljubljani, Fak. za računalništvo in informatiko
Pages:	1 spletni vir (1 datoteka PDF (IV, 102 str.))
ID:	25078342