Poboljšanje korisničkog iskustva uspostavom ANN predikcije i NLP chatbota

master's thesis

• Author: Kristina Puhanić
• Mentor: Diana Bratić (mentor)
• Committee member: Irena Bates (predsjednik povjerenstva)
• Committee member: Diana Bratić (član povjerenstva)
• Committee member: Suzana Pasanec Preprotić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Graphic Arts; Zagreb
• Defense date and country: 2022-09-29, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Graphic Technology

Abstract

Kako se komunikacija između davatelja usluge i korisnika povećava svakim danom zbog
velike ponude i potražnje za uslugom, primjena umjetne inteligencije u poslovne sektore
uvelike može olakšati i poboljšati karakteristike poput brzine i efikasnosti komunikacije
putem implementacije NLP chatbota koja će biti prezentirana u ovom diplomskom radu. Cilj
rada bio je ispitati posljedicu implementacije NLP chatbota i pouzdanosti ANN predikcije u
odnosu na korisničko iskustvo i ustanoviti hoće li izrađena aplikacija NLP chatbota
zadovoljiti snalažljivost u razumijevanju komunikacije s obzirom na pouzdanost algoritma.
Svrha leži i u provjeri uočavaju li se komparativnom analizom razlike u poboljšanju
korisničkog iskustva primjenom NLP chatbota i kakav utjecaj donosi brzina rada NLP
chatbota na korisničko iskustvo u odnosu na chatbot koji ne koristi NLP funkciju.
Programiranje aplikacije i simulacija su napravljeni pomoću programskog jezika Python u
integriranom razvojnom okruženju PyCharm koristeći se pritom modelom umjetne
neuronske mreže, te uporabom biblioteka kao što su PyTorch, NLTK i Tkinter. Izrada
eksperimentalnog dijela rada sastojala se od izrade aplikacije NLP chatbota, izrade
jednostavnog chatbota prikazanog putem simulativnog prikaza, te izrade komparativne
analize između NLP chatbota i chatbota utemeljenog na pravilima.
Predloženi model, prvenstveno sam koncept mogu poslužiti i za druga područja koja bi
značajno mogla profitirati od implementacije NLP chatbota, ne samo ona na kojima je
naglasak ležao u ovom radu.

Abstract (english)

As communication between service providers and users is increasing every day due to the
large supply and demand for services, the application of artificial intelligence in business
domains can greatly facilitate and improve characteristics such as the speed and efficiency
of communication through the implementation of NLP chatbots presented in this work. The
aim of this work is to investigate the consequences of NLP chatbot implementation and the
reliability of ANN prediction in terms of user experience, and to determine whether the
developed NLP chatbot application meets the ingenuity in understanding communication in
terms of algorithm reliability. Also, to verify whether differences in the improvement of user
experience by NLP chatbots can be observed through a comparative analysis and what impact
the speed of NLP chatbots has on the user experience compared to a chatbot without NLP
function.
The programming of the application and the simulation were done with the Python
programming language in the PyCharm integrated development environment, using the
artificial neural network model and libraries such as PyTorch, NLTK and Tkinter. The
creation of the experimental part consisted of the creation of an NLP chatbot application, the
creation of a simple chatbot represented by a simulation display, and the creation of a
comparative analysis between an NLP chatbot and a rule-based chatbot.
The proposed model, primarily the concept itself, can also be used for other areas that could
benefit significantly from the implementation of NLP chatbots, and not only for the areas
highlighted in this paper.

• Keywords: NLP; chatbot; AI; korisničko iskustvo; Pytorch
• Keywords (english): NLP; chatbot; AI; user experience; Pytorch
• Language: croatian
• URN:NBN: urn:nbn:hr:216:429009
• Study programme: Title: Graphic technology; specializations in: Graphic technology, Design of graphic products; Course: Design of graphic products; Study programme type: university; Study level: graduate; Academic / professional title: sveučilišni/a magistar/magistra inženjer/inženjerka grafičke tehnologije (sveučilišni/a magistar/magistra inženjer/inženjerka grafičke tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2024-07-08 09:21:48