Research shows that AI-driven chatbots have the potential to effectively support the management of stress-related mental health issues by offering easily accessible, resource-efficient, and non-stigmatized therapy. To optimize their effectiveness and usability, a deeper understanding of users' needs and preferences is required, along with a design tailored to these factors.  This study explores the use of AI-driven chatbots to address study-related stress among university students and identifies a research gap by integrating research on the therapeutic use of chatbots, stress management, and students' specific needs. The research question "How should an AI-driven chatbot be designed to assist students in managing study-related stress?" is used to investigate how to optimize chatbots to effectively support student well-being. By applying thematic analysis to 11 semi- structured interviews, the study examines the needs and preferences of Swedish university students and identifies key design features for this type of service.  The results indicate that students prefer the chatbot's role as a planning tool over therapeutic support, which differs from previous research. The findings also show that users prefer a communication style that is straightforward and focused on practical functions, and that the chatbot should be able to assume different roles, such as friend, study buddy, and mentor. Transparency regarding data protection and privacy is highlighted as a crucial factor for users' trust in the chatbot. The study contributes to the development of effective support systems to facilitate the university study environment by taking into account students' needs and preferences. / Forskning visar att AI-drivna chatbots har potential att effektivt stödja hanteringen avstressrelaterade psykiska problem genom att erbjuda lättillgänglig, resurseffektiv ochicke-stigmatiserad terapi. För att optimera deras effektivitet och användbarhet krävs endjupare förståelse för användarnas behov, samt en design som är anpassad efter dennaförståelse.Denna studie utforskar användningen av AI-drivna chatbots för att hanterastudierelaterad stress bland universitetsstudenter och identifierar en forskningsluckagenom att integrera forskning om terapeutisk användning av chatbots, stresshanteringoch studenters specifika behov. Forskningsfrågan "Hur bör en AI-driven chatbotdesignas för att hjälpa studenter att hantera studierelaterad stress?" används för attundersöka hur man kan optimera chatbots för att effektivt stödja studentersvälbefinnande. Genom att tillämpa tematisk analys på 11 semistrukturerade intervjuerundersöker studien behov hos svenska universitetsstudenter, samt identifierar viktigadesignegenskaper för denna typ av tjänst.Resultaten visar att studenter föredrar chatbotens roll som ett planeringsverktyg framförterapeutiskt stöd, vilket skiljer sig från tidigare forskning. Resultaten visar även attanvändarna föredrar en kommunikationsstil som är rak och fokuserad på praktiskafunktioner samt att chatboten bör kunna anta olika roller, såsom vän, studiekamrat ochmentor. Transparens kring dataskydd och integritet lyfts fram som en viktig faktor föranvändarnas förtroende för chatboten. Studien bidrar till utvecklingen av effektivastödsystem för att underlätta studiemiljön för universitetsstudenter genom att ta hänsyntill deras behov.

As research in Artificial Intelligence (AI) advances, it is crucial to focus on having seamless communication between humans and machines in order to effectively accomplish tasks. Smooth human-machine communication requires the machine to be sensible and human-like while interacting with humans, while simultaneously being capable of extracting the maximum information it needs to accomplish the desired task. Since a lot of the tasks required to be solved by machines today involve the understanding of images, training machines to have human-like and effective image-grounded conversations with humans is one important step towards achieving this goal. Although we now have agents that can answer questions asked for images, they are prone to failure from confusing input, and cannot ask clarification questions, in turn, to extract the desired information from humans. Hence, as a first step, we direct our efforts towards making Visual Question Answering agents human-like by making them resilient to confusing inputs that otherwise do not confuse humans. Not only is it crucial for a machine to answer questions reasonably, it should also know how to ask questions sequentially to extract the desired information it needs from a human. Hence, we introduce a novel game called the Visual 20 Questions Game, where a machine tries to figure out a secret image a human has picked by having a natural language conversation with the human. Using deep learning techniques like recurrent neural networks and sequence-to-sequence learning, we demonstrate scalable and reasonable performances on both the tasks. / Master of Science / Research in Artificial Intelligence has reached to a point where computers can answer natural freeform questions asked to arbitrary images in a somewhat reasonable manner. These machines are called Visual Question Answering agents. However, they are prone to failure from even a slightly confusing input. For example, for an obviously irrelevant question asked to an image, they would answer something non-sensical instead of recognizing that the question is irrelevant. Furthermore, they also cannot ask questions in turn to humans for clarification or for more information. These shortcomings not only harm their efficacy, but also harm their perceived trust from human users. In order to remedy these problems, we first direct our efforts towards making Visual Question Answering agents capable of identifying an irrelevant question for an image. Next, we also try to train machines to be able to ask questions to extract more information from humans to make an informed decision. We do this by introducing a novel game called the Visual 20 Questions game, where a machine tries to figure out a secret image a human has picked by having a natural language conversation with the human. Deep learning techniques such as sequence-to-sequence learning using recurrent neural networks make it possible for machines to learn how to converse based on a series of conversational exchanges made between two humans. Techniques like reinforcement learning make it possible for machines to better themselves based on rewards it gets for accomplishing a task in a certain way. Using such algorithms, we demonstrate promise towards scalable and reasonable performances on both the tasks.

The rise of conversational generative AI chatbots such as ChatGPT and Gemini is revolutionizing online interactions. Previous research has identified five categories of uses and gratifications (U&G) for users engaging with these chatbots: information seeking, task efficiency, social interaction, entertainment, and personalization. Despite the wide range of use cases, most chatbots provide one-size-fits-all text-based interactions, neglecting user preferences. Recent advancements are progressively introducing interactive features that empower users to control their interactions, such as choosing a preferred conversational style. However, despite these improvements in the industry, the interactivity in gen AI chatbots remains underexplored. This thesis serves as a user-centric foundational study of user engagement with gen AI chatbots by understanding users’ context of use across the five U&G dimensions, analyzing the limitations of text-based interactions, and proposing practical suggestions for desired interactive features.

Date: 2024-05-30 Course: Bachelor thesis in Business Administration, 15 cr School of Business, Society and Engineering, Mälardalen University Authors: Jonathan Granlund & David Lundström Title: Unveiling the Impact of AI-Powered Chatbots on Customer Acceptance in Sweden Supervisor:Emre Yildiz Keywords: AI-Chatbots. Technology Acceptance Model. Artificial Intelligence in Sweden. Artificial Intelligence in Customer Relations Research question: What are the effects of AI-based chatbots on customer acceptance in Sweden, including usefulness, ease of use, and trust? Purpose: This study aims to ascertain the ways in which perceived usefulness, ease of use, trust, information, innovativeness, and personalization affect customers' acceptance of AI chatbots in Sweden. Method: To get information on participants' opinions and future intentions about AI-powered chatbots, 150 people were asked to complete a survey to which 114 responded. To verify the proposed connections between the identified parameters and customer acceptance, the data was examined. Conclusion: The results validate that the behavioral intention to use AI-based chatbots is positively influenced by perceived usefulness, innovativeness, and personalization. Meanwhile perceived ease of use, trust and information were proved to be non-significant. These findings suggest that increasing chatbots' perceived usefulness, innovativeness and customization can encourage more people to use them. The factors influencing the acceptance of AI chatbots are empirically demonstrated in this study, which can help organizations and developers improve customer acceptance and engagement in the Swedish market. The study also emphasizes how crucial it is to use efficient marketing techniques and unambiguous communication to educate consumers about the features and advantages of AI chatbots, which will eventually lead to increased acceptance rates in customer applications.

ChatGPT and generative AI technologies have infiltrated our learning spaces, and, as a result, schools may be changed forever. While some educators may seek to ban the use of chatbots, motivated by a fear of the rampant plagiarism the technology might invite, I, however, write this dissertation with the intent of finding uses for AI as a participant in the teaching and learning of literature in the secondary and post-secondary English classroom. In this dissertation, I examine a series of problems, issues, and ideas raised by AI, situated in specific relationalities among readers and literary texts (students, teachers, and myself functioning as my main sites of inquiry) by engaging in literature-based experiments. Through reflecting on my experiences and experimenting alongside teachers, students, and AI, I have found that the problems and opportunities introduced by AI are not-so-new: they’re a re-presentation of the familiar, repackaged and amplified. Though this dissertation has not lent itself to the discovery of a singular conclusion, I have found, rather, grounds for further experimentation and provocation. As I conclude this dissertation, I attempt to identify some ways that teachers of English can utilize AI not as a tool for providing knowledge and information for students, but to rather utilize it as a thought-provoking companion for the teaching of literature.

Background  In the realm of contemporary marketing the implementation of AI Chatbots has emerged as a pivotal subject shaping brand perception and satisfaction in micro-businesses. As businesses increasingly integrate artificial intelligence into their customer service operations, understanding the implications of AI chatbot implementation on brand perception and satisfaction in micro-businesses becomes paramount. Purpose  The purpose of this paper is to gain a further understanding of how AI Chatbots affects brand perception and satisfaction in micro-businesses. The answer to this research question will not only contribute to resolving the identified gap but also offer actionable insights and recommendations for micro-business owners and managers. Method  A qualitative research approach has been selected. The approach entails employing AI Chatbots in three different micro-businesses and conducting interviews with three potential clients of each micro-business. Conclusion  The study demonstrates that AI Chatbots significantly influence brand perception and brand satisfaction in micro-businesses, enhancing the brand image by presenting a modern and friendly persona. However, their impact on brand satisfaction varies across different types of businesses; low-effort businesses see noticeable improvements while high-effort ones, like plastic surgery services, do not experience substantial gains due to already high expectations.

Large Language Models (LLMs) have the potential to enhance learning among students. These tools can be used in chatbot systems allowing students to ask questions about course material, in particular when plugged with the so-called Retrieval Augmented Systems (RAGs). RAGs allow LLMs to access external knowledge, which improves tailored responses when used in a chatbot system. This thesis studies different RAGs through an experimentation approach where each RAG is constructed using different sets of parameters and tools, including small and large language models. We conclude by suggesting which of the RAGs best adapts to high school courses in Physics and undergraduate courses in Mathematics, such that the retrieval systems together with the LLMs are able to return the most relevant answers from provided course material. We conclude with two RAG-powered LLM with different configurations performing over 64% accuracy in physics and 66% in mathematics.

[pt] A oferta de serviços por meio de interfaces conversacionais, ou chatbots, tem se tornado cada vez mais popular, com aplicações que variam de aplicativos de bancos e reserva de bilheteria a consultas em um banco de dados. No entanto, dado a quantidade massiva de dados disponível em alguns domínios, o usuário pode ter dificuldade em formular as consultas e recuperar as informações desejadas. Esta dissertação tem como objetivo investigar e avaliar o uso de recomendações na busca de informações numa base de dados de filmes através de chatbots. Neste trabalho, implementamos um chatbot por meio do uso de frameworks e técnicas da área de processamento de linguagem natural (NLP - Natural Language Processing). Para o reconhecimento de entidades e intenções, utilizamos o framework RASA NLU. Para a identificação das relações entre essas entidades, utilizamos as redes Transformers. Além disso, propomos diferentes estratégias para recomendações feitas a partir da ontologia de domínio. Para avaliação deste trabalho, conduzimos um estudo com usuários para avaliar o impacto das recomendações no uso do chatbot e aceitação da tecnologia por meio de um questionário baseado no Technology Acceptance Model (TAM). Por fim, discutimos os resultados do estudo, suas limitações e oportunidades de futuras melhorias. / [en] The offer of services provided through conversational interfaces, or chatbots, has become increasingly popular, with applications that range from bank applications and ticket booking to database queries. However, given the massive amount of data available in some domains, the user may find it difficult to formulate queries and retrieve the desired information. This dissertation investigates and evaluates the use of the recommendations in the search for information on a movie database through a chatbot. In this work, we implement a chatbot with the use of frameworks and techniques from the area of natural language processing (NLP). For the recognition of entities and intents, we use the RASA NLU framework. For the identification of relations between those entities, we use the Transformers networks. In addition, we propose different strategies for the recommendation from the domain ontology. To evaluate this work, we have conducted an empirical study with volunteer users to assess the impact of the recommendations on chatbot use and the acceptance of the technology through a survey based on the Technology Acceptance Model (TAM). Lastly, we discuss the results of this study, its limitations, and avenues for future improvements.

Syfte: Syftet med denna studie har varit att utforska och skapa ökad förståelse hur konsumenternas köpbeslut och efterköpsbeteende kan förändras av att heminredningsföretag inom e-handeln använder Artificiell Intelligens (AI) samt vilka eventuella konsekvenser AI användandet har på konsumenternas förtroende för företagen.  Problembakgrund: Med tanke på e-handelns framfart och att AI ständigt utvecklas är det relevant att ämnet diskuteras. Förtroende för AI skiljer sig mellan de svenska medborgarna samtidigt som implementeringen av AI ökar hos företagen.   Metod: Studien behandlar hur AI påverkar köpbeslut, efterköpsbeteende och förtroende med utgångspunkt i en kvalitativ studie och med en induktiv forskningsansats. Semistrukturerade intervjuer med femton personer genomfördes för att kunna besvara forskningsfrågorna.  Slutsats: Utifrån studiens empiri och analys kunde slutsatser dras att AI påverkar konsumenternas köpbeslut, efterköpsbeteende och förtroende på olika sätt i vardera steg. Detta ger en indikation från studien att företag som använder AI bör se till att det är genomtänkt och välutvecklat. Slutsatsen är att AI i dagsläget inte används av företag på ett tillräckligt utvecklat sätt utifrån dess potentiella kompetens, därav fyller det inte sin fulla funktion och konsumentens förtroende kan antas påverkas mer negativt än positivt. Om AI istället används på ett mer genomtänkt och välutvecklat sätt påverkar det konsumenternas förtroende positivt. / Purpose: The purpose of this study has been to explore and create greater understanding of how consumers' purchase decisions and post-purchase behavior change as a result of interior design companies in e-commerce using Artificial Intelligence (AI) and what possible consequences AI use has on consumers' trust in the companies. Problem background: Considering the progress of e-commerce and that AI is constantly developing, makes the topic relevant to discuss. Trust in AI differs between Swedish citizens, while the implementation of AI is increasing among companies. Method: The study investigates how AI affects purchase decisions, post-purchase behavior and trust based on a qualitative study with an inductive research approach. Semi-structured interviews with fifteen people were conducted in order to answer the research questions. Conclusion: Based on the study's empirical data and analysis, conclusions could be drawn that AI affects consumers' purchase decisions, post-purchase behavior and trust, in different ways at each stage. This gives an indication from the study that companies using AI should ensure that it is well thought out and well developed. The conclusion is that AI is currently not used by companies in a sufficiently developed way based on its potential competence, therefore it does not fulfill its full function and consumer trust can be assumed to be affected more negatively than positively. Instead, customer trust is enhanced when AI is applied in a well thought out and well developed manner.

Background: The increasing awareness of digitalization and specifically the emergence of Artificial Intelligence (AI) has made it possible for companies to apply chatbots. With the enhanced use of chatbots in digital settings, companies have applied chatbots to their digital customer service.  Purpose: This bachelor thesis aimed to explore chatbots' influence on customer satisfaction with digital customer service.  Method: The used research method for the study was qualitative research. To collect data for the research, the methods used were two focus groups. There were 12 participants, Swedish-speaking students from Linnaeus University, in Generation Z, both males and females.  Results: The results from the focus groups indicated that the chatbots' different characteristics and performance influenced the participants in variance. Half of the participants indicated the personal chatbot with friendly interaction influenced their customer satisfaction, and the other half influenced the impersonal chatbot. The participants agreed that it also depends on what situation the digital matter considers. All the participants agreed that the chatbots performance in general of response time and availability influence customer satisfaction.  Findings: Chatbots that are personal with friendly interactions influence customer satisfaction, for customers with complex digital matters. The second finding indicates that chatbots that are impersonal with intelligent interaction influence customer satisfaction, for customers with easy digital matters. The last findings indicates that, general performance of a chatbot, in terms of time-efficiency and availability, influences customer satisfaction for all digital matters.