DIGITAL LIBRARY
PERFORMANCE ANALYSIS OF VIRTUAL ASSISTANTS - A CASE STUDY BASED ON THE CROATIAN LANGUAGE
University of Rijeka, Faculty of Informatics and Digital Technology (CROATIA)
About this paper:
Appears in: INTED2024 Proceedings
Publication year: 2024
Pages: 7305-7315
ISBN: 978-84-09-59215-9
ISSN: 2340-1079
doi: 10.21125/inted.2024.1917
Conference name: 18th International Technology, Education and Development Conference
Dates: 4-6 March, 2024
Location: Valencia, Spain
Abstract:
A virtual assistant is a software agent that is able to perform various tasks or provide services to users, responding to user input, which may include commands, questions or verbal interactions.

A virtual assistant is a tool that is constantly evolving to help the user find the information they need and complete specific tasks. Virtual assistants have a wide range of applications: they automate certain phases of customer service, process simple requests and forward more complex requests to human agents. They help with everyday tasks such as planning appointments, sending messages or reminders, searching for relevant information on the internet, providing weather forecasts, news, or other useful data, managing smart home devices, setting reminders, making calls or sending messages. They can be integrated into company systems to facilitate internal processes and communication, and much more. The possibilities and use of virtual assistants are growing rapidly due to advances in technologies such as natural language processing, machine learning and artificial intelligence. As a result, major technology companies have developed their own versions of virtual assistants, such as Apple's Siri, Samsung's Bixby, Microsoft's Cortana, Google's Google Assistant, etc. Virtual assistants are often used to automate customer support via email, chat, or social media. They serve as the first point of contact for a company's customer service and forward complex cases to human agents if necessary. Virtual assistants enable round-the-clock customer support with automated interactions and eliminate repetitive processes.

Virtual assistants can also be used in education to support learning and teaching. For example, teachers and students can use them for quick internet searches, to set reminders for important tasks, homework, or exams, or to record oral instructions or notes from teachers or students. For foreign students, they can help with translation and facilitate two-way communication, etc. By customizing the virtual assistant to meet specific educational needs and goals, learning efficiency and teacher support can be improved.

Virtual assistants and chatbots are often used interchangeably, but the difference between virtual assistants and chatbots often lies in the level of complexity and customizability. While chatbots are often programmed for specific tasks or scenarios and work based on a predefined script, virtual assistants are able to understand the context and adapt to the user experience (UX). They are able to handle more complex requests, understand contextual information from previous conversations and provide more personalized and useful answers.

The article briefly describes selected virtual assistants (Siri, Bixby and Google Assistant), their strengths and challenges. The success of the selected virtual assistants in accomplishing tasks was compared in several categories (call category, message and email category, settings management category, calculation category, reminder category, navigation category, search category, entertainment category, and third-party applications category). Each virtual assistant was given a score from 1 to 5 based on their performance in each category. The tasks were conducted in both English and Croatian, including Croatian names and terms, to examine how well virtual assistants adapt to languages other than English.
Keywords:
Virtual assistant, Siri, Bixby, Google Assistant, examples.