Coherence is the quality that gives a text its conceptual unity, making a text a coordinated set of connected parts rather than a random group of sentences (turns, in the case of dialogue). Hence, coherence is an integral property of human communication, necessary for a meaningful discourse both in text and dialogue. As such, coherence can be regarded as a requirement for conversational agents, i.e. machines designed to converse with humans. Though recently there has been a proliferation in the usage and popularity of conversational agents, dialogue coherence is still a relatively neglected area of research, and coherence across multiple turns of a dialogue remains an open challenge for current conversational AI research. As conversational agents progress from being able to handle a single application domain to multiple ones through any domain (open-domain), the range of possible dialogue paths increases, and thus the problem of maintaining multi-turn coherence becomes especially critical. In this thesis, we investigate two aspects of coherence in dialogue and how they can be used to design modules for an open-domain coherent conversational agent. In particular, our approach focuses on modeling intentional and thematic information patterns of distribution as proxies for a coherent discourse in open-domain dialogue. While for modeling intentional information we employ Dialogue Acts (DA) theory (Bunt, 2009); for modeling thematic information we rely on open-domain entities (Barzilay and Lapata, 2008). We find that DAs and entities play a fundamental role in modelling dialogue coherence both independently and jointly, and that they can be used to model different components of an open-domain conversational agent architecture, such as Spoken Language Understanding, Dialogue Management, Natural Language Generation, and open-domain dialogue evaluation. The main contributions of this thesis are: (I) we present an open-domain modular conversational agent architecture based on entity and DA structures designed for coherence and engagement; (II) we propose a methodology for training an open-domain DA tagger compliant with the ISO 24617-2 standard (Bunt et al., 2012) combining multiple resources; (III) we propose different models, and a corpus, for predicting open-domain dialogue coherence using DA and entity information trained with weakly supervised techniques, first at the conversation level and then at the turn level; (IV) we present supervised approaches for automatic evaluation of open-domain conversation exploiting DA and entity information, both at the conversation level and at the turn level; (V) we present experiments with Natural Language Generation models that generate text from Meaning Representation structures composed of DAs and slots for an open-domain setting.

This thesis explores the use of advanced language models, specifically OpenAI’s Generative Pretrained Transformer (GPT), in the context of interactive tutoring systems built within a Unity-based game environment. The central problem addressed is whether the recent advancements in large language models make them feasible and useful to function as tutors specifically in providing meaningful, engaging, and educationally rich user interactions on a dialogue based learning platform developed by Fictive Reality. There is also a comparison on the effectiveness of GPT versus the model that previously powered the learning platform built in Rasa. The importance of this problem lies in offering people learning opportunities that might not otherwise be available to them, and in seeing if recent advancements in generative AI are sufficient for developing useful interactive AI tutors of soft skills. The Fictive Reality learning platform is powered by a Rasa model that generates appropriate responses to users in the context of roleplay-based learning scenarios while keeping an internal state of the progress of the dialogue. The project entails replacing this model with GPT and a comparison of their performance and respective merits. We also explored the potential for a hybrid model, leveraging the strengths of both systems. Using Rasa for internal state tracking and answering simpler queries, and utilizing GPT to handle those queries whose intent Rasa cannot determine. The first part of this project was integrating GPT with the existing functionality of the platform, this includes changes to the platform that allow people to create and play GPT powered learning scenarios and adopting the existing features and user interface. Additionally, prompt engineering GPT to act as a tutor and to stay within the context of a learning environment. Changes had to be made to the platform so that the already existing features of Rasa scenarios could be replicated in GPT scenarios. Finally there is a systematic comparison of the user experience and performance metrics when interacting with either a GPT or a Rasa chatbot in a learning scenario. Specifically these metrics are determined from the conversational flow between bot and user, the context and continuity, finish rate, chit-chat handling and length of average session. The results suggest a distinct user preference for the GPT model due to its superior conversational capabilities, despite Rasa’s faster response times and state-tracking feature. The study suggest that GPT is sufficient for creating useful learning scenarios in restricted contexts. Therefore we suggest that large language models can be leveraged in interactive learning systems, with potential impacts on edtech, AI in education, and conversational AI. / Detta examensarbete utforskar användningen av avancerade språkmodeller, särskilt OpenAI’s Generative Pretrained Transformer (GPT), tillsammans med interaktiva handledningssystem byggda i en Unity-baserad spelmiljö. Det centrala problemet som tas upp är om det är genomförbart och användbart att använda GPT som handledare. Vidare genomfördes också en jämförelse av effektiviteten hos GPT jämfört med en mer traditionell modell, Rasa, när det gäller att tillhandahålla meningsfulla, engagerande och lärorika interaktioner. Detta problem har betydelse för att erbjuda människor lärandemöjligheter som annars kanske inte skulle vara tillgängliga för dem och för att se om de senaste framstegen inom generativ AI är tillräckliga för användbar interaktiv handledning av mjuka färdigheter, så kallade soft skills". Lärplattformen Fictive Reality drivs av en Rasa-modell som genererar lämpliga svar till användare i samband med vissa inlärningsscenarier samtidigt som man behåller ett internt tillstånd av dialogens framsteg. Projektet syftar till att ersätta denna modell med GPT och göra en jämförelse av prestandan och hos respektive modell. Vi undersökte också potentialen för en hybridmodell som utnyttjar båda systemens styrkor genom att använda Rasa för intern tillståndsspårning och svara på enklare frågor, och använda GPT för att hantera de frågor vars avsikt Rasa inte kan avgöra. Den första delen av projektet var att integrera GPT med plattformens befintliga funktionalitet, detta inkluderar förändringar av plattformen som gör det möjligt för människor att skapa och spela GPT-drivna inlärningsscenarier med det befintliga användargränssnittet och funktioner för Rasa-drivna scenarier. Förändringar var tvungna att göras på plattformen så att de redan befintliga funktionerna i Rasa-scenarier kunde replikeras i GPT-scenarier. Slutligen gjordes en systematisk jämförelse av prestandan och användarupplevelsen när man interagerar med antingen en GPT- eller en Rasa-chatbot i ett inlärningsscenario. Resultaten tyder på en distinkt användarpreferens för GPT-modellen på grund av dess överlägsna konversationsförmåga, trots Rasa:s snabbare svarstider och tillståndsspårningsfunktion. Studien tyder på att GPT är tillräckligt för att skapa användbara lärande scenarier i begränsade sammanhang. Denna studie tyder på att stora språkmodeller kan utnyttjas i interaktiva inlärningssystem, med potentiella effekter på edtech, AI inom utbildning och konversations-AI-områden.

The success of software development projects heavily depends on effectively capturing and meeting stakeholders' requirements. However, involving multiple stakeholders with diverse backgrounds and objectives often leads to conflicts among these requirements. These conflicts represent inconsistencies in the system design, resulting in various challenges, including project delays, increased costs, and potential system failures. Previous research has primarily focused on identifying conflicts with algorithms or negotiation, while conversational AI's potential to detect conflicts in real-time has been neglected. This thesis study addresses the challenge of requirement conflicts by proposing a novel approach that leverages conversational AI in the form of a chatbot. The chatbot, developed using the Rasa platform, enables real-time detection of conflicts, focusing on three general types: duplicated (similar), incompatible, and contradictory requirements. During the study, the design science research method is employed to guide the chatbot's development. Further, an experiment is applied to evaluate the chatbot's performance compared with domain experts using four different datasets. The experiment results are presented using F1 scores, which calculate precision and recall for both the chatbot and the experts on each dataset. Overall, the chatbot scored 0.8, while the experts achieved a slightly higher score of 0.86. To determine if there was a statistically significant difference between the two performances, a Wilcoxon signed-rank test was conducted on the results. The analysis showed no significant difference in the F1 score between the chatbot and the experts, indicating the chatbot's feasibility and effectiveness in detecting conflicts. The contribution of this thesis study can advance requirements engineering by providing a user-friendly and efficient method for real-time conflict detection, enhancing the quality and overall success of software development projects.

Humanizing Technologies GmbH develops Intelligent software for the humanoid robots from Softbank Robotics. The main objective of this thesis is to develop and deploy Conversational Artificial Intelligence software into the humanoid robots using deep learning techniques. Development of conversational agents using Machine Learning or Artificial Intelligence is an intriguing issue with regards to Natural Language Processing. Great research and experimentation is being conducted in this area. Currently most of the chatbots are developed with rule based programming that cannot hold conversation which replicates real human interaction. This issue is addressed in this thesis with the development of Deep learning conversational AI based on Sequence to sequence, Attention mechanism, Transfer learning, Active learning and Beam search decoding which emulates human like conversation. The complete end to end conversational AI software is designed, implemented and deployed in this thesis work according to the conceptual specifications. The research objectives are successfully accomplished and results of the proposed concept are dis- cussed in detail.

Bakgrund: AI, specifikt konversations-AI som OpenAI:s ChatGPT, växer snabbt i både privata och professionella sammanhang, vilket erbjuder möjligheter till kostnadsbesparingar och modernisering för företag. ChatGPT kan simulera mänskliga konversationer, vilket kan ge fördelar i flera olika industrier och kan genom samarbete mellan människa och AI potentiellt förbättra anställdas produktivitet. Det huvudsakliga forskningsproblemet är att identifiera faktorer som påverkar systemutvecklarens användning av ChatGPT och beakta dess design och implementation för att minska potentiella negativa effekter. Syfte:  Denna studie syftar till att undersöka de faktorer som påverkar användares adoption ChatGPT som ett verktyg för att stödja systemutvecklare. Dessutom syftar studien till att identifiera hur ChatGPT kan hjälpa systemutvecklare i deras dagliga arbete och vilka hinder som finns för att inkorporera ChatGPT i denna kontext. Metod: Genom en fallstudieansats med kvalitativa och kvantitativa datainsamlingsmetoder, använder studien positivistiska och interpretivistiska paradigm. Resultat: Resultatet visar att den uppfattade förmågan hos ChatGPT att förbättra effektiviteten och generera korrekta svar påverkar avsikten att använda tekniken. Faktorer som tidsbesparing, produktivitetsförbättring och användarvänlighet gav dock inte statistiskt signifikanta resultat. Utvecklare finner ChatGPT användbart för att förenkla uppgifter och hjälpa juniora utvecklare, men det finns oro för att hantera komplexa uppgifter och säkerhetsfrågor. Slutsatser: Användarnas acceptans av ChatGPT drivs främst av den uppfattade precisionen och effektiviteten. ChatGPT kan hjälpa till med uppgifter som felsökning, kodgenerering, kodrefaktorering, kodoptimering och teknisk dokumentation, men med vissa potentiella begränsningar när det gäller hantering av alltför komplex kod. Trots detta så finns hinder för införandet i form av oro för integritet, säkerhet och brist på medvetenhet samt funktionella begränsningar. Följder: De insikter som vunnits kan indirekt gynna företag, inklusive vår affärspartner CGI, genom att bidra till beslutsfattandeprocesser relaterade till adoption och användning av ChatGPT. / Background: AI, specifically conversational AI like OpenAI's ChatGPT, is rapidly expanding in personal and professional settings, offering cost-cutting and modernization opportunities for businesses. This technology, capable of simulating human-like conversations, holds promise across various industries, potentially enhancing productivity through human-AI collaboration. The main research problem is to identify factors influencing system developers' adoption of ChatGPT, considering its design and implementation to mitigate potential negative impacts. Aim: This study aims to investigate the factors that influence user adoption of ChatGPT as a tool to support system developers. Additionally, it aims to identify how ChatGPT can aid system developers in their daily work, and challenges associated with incorporating ChatGPT in this context. Method: Using a case study approach with qualitative and quantitative data collection methods, the study employs positivist and interpretivist philosophical paradigms. Results: Results showed that the perceived ability of ChatGPT to enhance efficiency and generate accurate responses significantly impacts adoption intentions. When examining aspects related to timesaving, productivity enhancement, and user-friendliness, no statistically significant results were found. Among developers, ChatGPT is considered valuable for simplifying tasks and assisting junior developers. There are concerns regarding its capability to handle complex tasks and potential security issues. Suggestions for improvement include better integration with integrated development environments (IDEs) and enhanced accuracy. Conclusions: The findings highlight perceived accuracy and efficiency as driving factors for user adoption regarding ChatGPT. ChatGPT can support tasks like debugging, code generation, code refactoring, code optimization, and technical documentation. However, there may be some potential limitations when dealing with overly complex code. Barriers to adoption include concerns about integrity and security, lack of awareness, and functional limitations. Implications: The insights gained can indirectly benefit companies, including our business partner CGI, by guiding decision-making processes related to the effective adoption and utilization of ChatGPT.

This paper evaluates and analyzes three conversational AI-platforms; Dialogflow (Google), Watson Assistant (IBM) and Teneo (Artificial Solutions) on how they perform based on a set of criteria; pricing model, ease-of-use, efficiency, experience working in the software and what results to expect from each platform. The main focus was to investigate the platforms in order to acquire an understanding of which platform would best be suited for enterprises. The platforms were compared by performing a variety of tasks aiming to answer these questions. The technical research was combined with an analysis of each company’s pricing model and strategy to get an understanding of how they target their products on the market. This study concludes that different softwares may be suitable for different settings depending on the size of an enterprise and the demand for complex solutions. Overall, Teneo outperformed its competitors in these tests and seems to be the most scalable solution with the ability to create both simple and complicated solutions. It was more demanding to get started in comparison with the other platforms, but became more efficient as time progressed. Some findings include that Dialogflow and Watson Assistant lacked capabilities when faced with  complex and complicated tasks. From a pricing strategy point of view, the companies are similar in their approach but Artificial Solutions and IBM has more flexible methods while Google has a fixed pricing strategy. Combining the pricing strategy and technical analysis this implicates that Teneo would be a better choice for larger enterprises while Watson Assistant and Dialogflow may be more suitable for smaller ones. / Det här arbetet evaluerar och analyserar tre konversationella AI-plattformar; Dialogflow (Google), Watson Assistant (IBM) och Teneo (Artificial Solutions) utifrån hur de presterar baserat på ett antal  kriterier; prismodell, enkel användning, effektivitet, upplevelse att arbeta i programvaran och vilka resultat man förväntar sig från varje plattform. Huvudsakligt fokus var att undersöka plattformarna för att få en uppfattning om vilken plattform som skulle passa bäst för företag. Plattformarna jämfördes genom att utföra en mängd olika uppgifter som syftade till att besvara dessa frågor. Den tekniska forskningen kombinerades med en analys av varje företags prismodell och prisstrategi för att få en uppfattning av hur de riktar sina produkter på marknaden. Denna studie drar slutsatsen att olika programvaror kan vara lämpliga för olika sammanhang beroende på ett företags storlek och dess efterfrågan på komplexa lösningar. Sammantaget överträffade Teneo sina konkurrenter i dessa tester och verkar vara den mest skalbara lösningen med förmågan att skapa både enkla och komplicerade lösningar. Det var mer krävande att komma igång i jämförelse med de andra plattformarna, men det blev mer effektivt med tiden. Vissa fynd inkluderar att Dialogflow och Watson Assistant saknade kapacitet när de mötte komplexa och komplicerade uppgifter. Från en prissättningsstrategisk synvinkel är företagen liknande i sin metod men Artificial Solutions och IBM har mer flexibla metoder medan Google har en fast prissättningstrategi. Genom att kombinera prisstrategi och teknisk analys innebär detta att Teneo skulle vara ett bättre val för större företag medan Watson Assistant och Dialogflow kan vara mer lämpade för mindre.

Användningen av artificiell intelligens (AI) inom hälso- och sjukvården har potential att minska den administrativa bördan och förbättra patientvården genom att effektivisera dokumentationsprocesser. Trots dessa fördelar kvarstår utmaningar som bristande integration och tung arbetsbelastning för vårdpersonal. Denna studie undersöker hur AI, särskilt konversations AI (CAI), kan användas för att förbättra dokumentationen i vården och stödja vårdpersonalens dagliga arbete. Syftet är att identifiera viktiga faktorer som påverkar användarnas upplevelse samt hur de bör utformas för att stötta vårdpersonalen vid dokumentation med hjälp av CAI För att uppnå detta genomfördes semistrukturerade intervjuer med vårdpersonal, kompletterat med konceptbilder för att stimulera diskussioner och förståelse. Data analyserades med både deduktiva och induktiva analysmetoder. Resultaten visar att viktiga faktorer för utformning av CAI-system inkluderar anpassningsbara konversationsgränssnitt, möjligheter för användaren att ge feedback och transparenta förklaringar av systemets beslut. Dessutom framhävdes vikten av att CAI-system kan integreras sömlöst med befintligaarbetsflöden och att de kan anpassa sig efter individuella användarbehov. Andra viktiga aspekter är att systemet ska vara intuitivt, tillförlitligt och stödja en effektiv arbetsmiljö. Studien bekräftar att flera faktorer är avgörande för att utforma effektiva förklaringar iAI-system, men ytterligare forskning behövs för att utforska hur dessa riktlinjer påverkar tillit och förståelse över tid. / The use of artificial intelligence (AI) in healthcare has the potential to reduce administrative burdens and improve patient care by streamlining documentation processes. Despite these benefits, challenges such as lack of integration and heavy workloads for healthcare staff persist. This study explores how AI, particularly conversational AI (CAI) can be utilized to enhance healthcare documentation and support the daily tasks of healthcare professionals. The aim is to identify key factors that are important for user experience and how CAI-systems should be designed to support healthcare staff in documentation. To achieve this, semi-structured interviews with healthcare professionals were conducted, supplemented with concept images to stimulate discussions and understanding. Data were analyzed using both deductive and inductive methods. The results indicate that key factors for designing CAI-systems include customizable converational interfaces, opportunities for user feedback, and transparent explanations of the system's decisions. Additionally, the importance of seamless integration of CAI-systems with existing workflows and their adaptability to individual user needs was emphasized. Other critical aspects are that the system should be intuitive, reliable, and support an efficient work environment. The study confirms that several factors are crucial for designing effective CAI-systems in healthcare. However, further research is needed to explore how these guidelines affect trust and understanding over time.

The research of open-domain, knowledge-grounded dialogue systems has been advancing rapidly due to the paradigm shift introduced by large language models (LLMs). While the strides have improved the performance of the dialogue systems, the scope is mostly monolingual and English-centric. The lack of multilingual in-task dialogue data further discourages research in this direction. This thesis explores the use of transfer learning techniques to extend the English-centric dialogue systems to multiple languages. In particular, this work focuses on five typologically diverse languages, of which well-performing models could generalize to the languages that are part of the language family as the target languages, hence widening the accessibility of the systems to speakers of various languages. I propose two approaches: Multilingual Retrieval-Augmented Dialogue Model (xRAD) and Multilingual Generative Dialogue Model (xGenD). xRAD is adopted from a pre-trained multilingual question answering (QA) system and comprises a neural retriever and a multilingual generation model. Prior to the response generation, the retriever fetches relevant knowledge and conditions the retrievals to the generator as part of the dialogue context. This approach can incorporate knowledge into conversational agents, thus improving the factual accuracy of a dialogue model. In addition, xRAD has advantages over xGenD because of its modularity, which allows the fusion of QA and dialogue systems so long as appropriate pre-trained models are employed. On the other hand, xGenD takes advantage of an existing English dialogue model and performs a zero-shot cross-lingual transfer by training sequentially on English dialogue and multilingual QA datasets. Both automated and human evaluation were carried out to measure the models' performance against the machine translation baseline. The result showed that xRAD outperformed xGenD significantly and surpassed the baseline in most metrics, particularly in terms of relevance and engagingness. Whilst xRAD performance was promising to some extent, a detailed analysis revealed that the generated responses were not actually grounded in the retrieved paragraphs. Suggestions were offered to mitigate the issue, which hopefully could lead to significant progress of multilingual knowledge-grounded dialogue systems in the future.

Work email is one of the organisations most critical tool today. It`s have become a standard way to communicate internally and externally. It can also affect our well-being. Email overload has become a well-known issue for many people. With interviews, follow up interviews, and a workshop, three persons from an accounting firm prioritise pre-define emails needs. And identified several other email needs that were added to the priority list. A thematic analysis and summarizing of a Likert scale was conducted to identify underlying work email needs and work email needs that are not apparent. Three work email needs were selected and using scenario-based methods and the elements of PACT to investigating how the characteristics of a chatbot can help solve the identified work email overload issue? The result shows that email overload is percept different from individual to individual. The choice of how email is handled and email activities indicate how email overload feeling is experienced. The result shows a need to get a sense of the email content quickly, fast collect financial information and information from Swedish authorities, and repetitive, time-consuming tasks. Suggestions on how this problem can be solved have been put forward for many years, and how to use machine learning to help reduce email overload. However, many of these proposed solutions have not yet been implemented on a full scale. One conclusion may be that since email overload is not experienced in the same way, individuals have different needs - One solution does not fit all. With the help of the character of a chatbot, many problems can be solved. And with a technological character of a chatbot that can learn individuals' email patterns, suggest email task to the user and performing tasks to reducing the email overload perception. Using keyword for email intents to get a sense of the email content faster and produce quick links where to find information about the identified subject. And to work preventive give the user remainder and perform repetitive tasks on specific dates.

No description available.