Use Of Conversational Agents Research Articles

Examining the persuasiveness of text and voice agents: prosody aligned with information structure increases human-likeness, perceived personalisation and brand attitude

ABSTRACT To give product and brand recommendations, marketers make use of conversational agents which increasingly communicate via voice rather than text. Existing research comparing the persuasiveness of text and voice agents showed mixed results. The quality of the speech synthesis employed may strongly influence consumers’ responses. This study investigates to what extent a voice agent with pragmatically aligned prosody is more persuasive (i.e. yields a more positive brand attitude) than an agent with a standard voice or text, and whether perceived human-likeness and perceived personalisation provide an underlying mechanism to explain these differences. In an experiment (n = 212), participants interacted with a conversational agent that recommended a camera. Results showed that a voice agent using prosody aligned to the information state of the user is more persuasive than a text agent. This effect is mediated by perceived human-likeness and perceived personalisation. Hence, aligned prosody can make synthetic speech meet a certain quality threshold to be perceived as more human-like. Theoretically, this study helps to unravel why conversational agents with human-like features are more persuasive.

Top-funded companies offering digital health interventions for the prevention and treatment of depression: a systematic market analysis

BackgroundDigital innovations can reduce the global burden of depression by facilitating timely and scalable interventions. In recent years, the number of commercial Digital Health Interventions for Depression (DHIDs) has been on the rise. However, there is limited knowledge on their content and underpinning scientific evidence. This study aimed to: (i) identify the top-funded companies offering DHIDs and (ii) provide an overview of their interventions, including scientific evidence, psychotherapeutic approaches and use of novel technologies.MethodsA systematic search was conducted using two venture capital databases to identify the top-30 funded companies offering DHIDs. In addition, studies related to the DHIDs’ were identified via academic databases and hand-searching. The methodological quality of the publications was evaluated using the Mixed Methods Appraisal Tool.ResultsThe top-30 funded companies offering DHIDs received a total funding of 2,592 million USD. Less than half of the companies produced any scientific research associated with their DHIDs, with a total of 83 publications identified. Twenty-five publications were randomised control trials, of which 15 reported moderate-to-large effects in reducing depression symptoms. Regarding novel technologies, few DHIDs incorporated the use of conversational agents or low-burden sensing technologies.ConclusionsFunding received by top-funded companies was not related to the amount of scientific evidence provided on their DHIDs. There was a strong variation in the quantity of evidence produced and an overall need for more rigorous effectiveness trials. Few DHIDs used automated approaches such as conversational agents, limiting their scalability.

Conversational Agent Utilization Patterns of Individuals with Autism Spectrum Disorder

AbstractIndividuals with autism spectrum disorder (ASD) face unique challenges in their social interactions. The use of conversational agents (CA) can provide support and help reduce barriers to care. However, research on CA’s by individuals with ASD is limited. The present study sought to better understand CA utilization patterns by users with ASD. A subset of data was collected from users of Wysa, a mental health CA. Engagement with the CA, utilization of offered mental health interventions (CA tools), collection of energy scores, depression (PHQ-9), and anxiety (GAD-7) outcomes were gathered and analyzed. Users engaged with Wysa on average 8.59 days and had a median of 97 conversational exchanges. Almost half of the users utilized at least one of the 230 tools offered. The most frequently used interventions focused on mindfulness, thought recording, sleep, grounding, and social support. Energy scores were reported on average 10.59 times, and the mean energy score was 42.77 out of 100. Mean baseline PHQ-9 and GAD-7 scores were 10.34 and 8.81, respectively. Overall, the current findings show that users with ASD engaged regularly with a CA, despite its targeted design for other mental health concerns such as depression and anxiety, rather than ASD. If users with ASD engage with these types of resources, this could become a new avenue of support to a population facing multiple challenges to accessing treatment.

Confirmatory Exploration of innovation in virtual exchanges through a studyon Virtual Agents

The use of Conversational Agents for consumer services continues to grow, and the service industry is no exception. Carried out among 340 Moroccan consumers, this study shows the importance of perceived anthropomorphism, social presence, perceived trust, social influence, perceived cost, and facilitating conditions of a conversational agent in increasing adoption intention using the UTAUT model. The expected effort had no effect. This research confirms the moderating role of age on the relationship between social influence and intention to use, while gender and level of education had no moderating effect. The technology acceptance model tested is a powerful tool for understanding why people do or do not use conversational agents. It can explain nearly 75,8% of the variance in behavioral intention. Various recommendations are put forward to optimize the design and successful implementation of chatbots, enabling them to take their rightful place among digital tools.

Mitigating Algorithm Aversion in Recruiting: A Study on Explainable AI for Conversational Agents

The use of conversational agents (CAs) based on artificial intelligence (AI) is becoming more common in the field of recruiting. Organizations are now adopting AI-based CAs for applicant (pre-)selection, but negative news coverage, especially the black-box character of AI, has hindered adoption. So far, little is known about the contextual factors influencing users' perception of AI-based CAs in general and the effect of provided explanations by explainable AI (XAI) in particular. While research on algorithm aversion provides some initial explanations, information regarding the effects of different XAI approaches on different types of decisions on the attitudes of (potential) applicants is scarce. Therefore, in this study, we use a quantitative, quota-representative study (n = 490) to assess the acceptance of CAs in recruiting. By applying an experimental within-subject design, we provide a more nuanced perspective on why and when providing explanations increases user acceptance. We also show that contextual factors such as the type of assessed skills are major determinants of this effect, and we conclude that XAI is not a "one-size-fits-all approach." Based on the insight that contextual factors of the decision problem are more important than the type of XAI approach itself, we argue that the use and the effects of explainability in recruiting need a more nuanced perspective, focusing on the fit of explanations with the user's characteristics and preferences.

Productivity of incident management with conversational bots-a review

<p>The use of conversational agents (bots) in information systems managed by company’s increases productivity in the development of activities focused on processes such as customer service, healthcare, and presentation. The present work is a systematic literature review that collects articles from 2019 to 2022 in the databases Scopus, Springer, Willey, Indexes-Csic, Taylor & Francis, Pubmed, and Ebsco Host. PRISMA methodology was used to systematize 47 relevant articles. As a result of the analysis, 2/19 very important benefits were obtained, which are: helping to obtain information and facilitating customer service; as for the types of conversational bots, a total of 9 types were found, of which conversational agents and chatbots with artificial intelligence (AI) are the most common; in the case of processes, 3/5 processes that optimize conversational bots were found, where the most prominent are: teaching process, health processes, and customer service processes. An architecture model for conversational bots in incident management is also proposed.</p>

How speaking versus writing to conversational agents shapes consumers’ choice and choice satisfaction

The use of conversational agents (e.g., chatbots) to simplify or aid consumers’ purchase decisions is on the rise. In designing those conversational agents, a key question for companies is whether and when it is advisable to enable voice-based rather than text-based interactions. Addressing this question, this study finds that matching consumers’ communication modality with product type (speaking about hedonic products; writing about utilitarian products) shapes consumers’ choice and increases choice satisfaction. Specifically, speaking fosters a feeling-based verbalizing focus, while writing triggers a reason-based focus. When this focus matches consumers’ mindset in evaluating the product type, preference fluency increases, thereby enhancing choice satisfaction. Accordingly, the authors provide insights into managing interactions with conversational agents more effectively to aid decision-making processes and increase choice satisfaction. Finally, they show that communication modality can serve as a strategic tool for low-equity brands to better compete with high-equity brands.

A systematic review A Conversational interface agent for the export business acceleration

Conversational agents, which understand, respond to, and learn from each interaction using Automatic Speech Recognition (ASR), Natural Language Processing (NLP), Advanced Dialog Management, and Machine Learning (ML), have become more common in recent years. Conversational agents, also referred to as chatbots, are used to have real-time conversations with individuals. As a result, conversational agents are now being used in a variety of sectors, including those in education, healthcare, marketing, customer assistance, and entertainment. Conversational agents, which are frequently used as chatbots and virtual or AI helpers, show how computational linguistics is used in everyday life. It can be challenging to pinpoint the variables that affect the use of conversational agents for business acceleration and to defend their utility in order to enhance export company. This paper provides a summary of the evolution of conversational agents from a straightforward model to a sophisticated intelligent system, as well as how they are applied in various practical contexts. This study contributes to the body of literature on information systems by contrasting the different conversational agent types based on the export business acceleration interface. This paper also identifies the challenges conversational applications experience today and makes recommendations for further research.

Hey Google, What is in the News? The Influence of Conversational Agents on Issue Salience

The emerging use of Conversational Agents (CAs), such as Google Assistant, highlights the role of algorithmic gatekeeping power in news consumption. However, our knowledge of the effects of CAs on shaping the public’s perception of the most important topics (issue salience) is limited. To investigate this, we conducted a seven-day longitudinal survey in the Netherlands, comparing the most important issues reported by users (N = 352) across various channels, including CAs, social media, and news websites. Our aim was to analyze how CAs influence the transfer of issue salience from the channel to the user. While our hypotheses were not supported, different patterns emerged. Users consuming news through CAs exhibited a have a lower probability of agreeing on the most important issues in the Netherlands compared to participants consuming information on social media and news websites. This suggests that news consumption through CAs may offer less shared context compared to other channels. Furthermore, trust in the channel and attention to the news were found to be significant predictors of agreement on issue salience. These findings raise important questions for further research, particularly regarding the impact and democratic role of CAs within the broader news ecosystem.

Conversational Agents and Avatars for Cardiometabolic Risk Factors and Lifestyle-Related Behaviors: Scoping Review.

In recent years, there has been a rise in the use of conversational agents for lifestyle medicine, in particular for weight-related behaviors and cardiometabolic risk factors. Little is known about the effectiveness and acceptability of and engagement with conversational and virtual agents as well as the applicability of these agents for metabolic syndrome risk factors such as an unhealthy dietary intake, physical inactivity, diabetes, and hypertension. This review aimed to get a greater understanding of the virtual agents that have been developed for cardiometabolic risk factors and to review their effectiveness. A systematic review of PubMed and MEDLINE was conducted to review conversational agents for cardiometabolic risk factors, including chatbots and embodied avatars. A total of 50 studies were identified. Overall, chatbots and avatars appear to have the potential to improve weight-related behaviors such as dietary intake and physical activity. There were limited studies on hypertension and diabetes. Patients seemed interested in using chatbots and avatars for modifying cardiometabolic risk factors, and adherence was acceptable across the studies, except for studies of virtual agents for diabetes. However, there is a need for randomized controlled trials to confirm this finding. As there were only a few clinical trials, more research is needed to confirm whether conversational coaches may assist with cardiovascular disease and diabetes, and physical activity. Conversational coaches may regulate cardiometabolic risk factors; however, quality trials are needed to expand the evidence base. A future chatbot could be tailored to metabolic syndrome specifically, targeting all the areas covered in the literature, which would be novel.

Conversational agents: a survey on culturally informed design practices

The development and use of conversational agents grow every day. However, those systems still need to meet users' expectations. As a result, new design practices for conversational agents are emerging. One facet is the culturally informed design of those systems. Tailoring conversational agents culturally can increase engagement, trustworthiness, and acceptance. While an emerging trend, literature still needs to discuss research and application of design practices related to culture and conversational agents. The present study investigates 5398 articles involving conversational agents and culture over the last seven years. In this context, we selected 23 articles based on their approaches and objectives related to our main subject. The findings indicate that most studies in the field are mainly focused on the feasibility of cultural markers tailored to those systems. The main contribution is the identification of the main challenges, contexts and design practices in the field. These results contributed to discussions regarding conversational agents and highlighted some research gaps for future study.

Using Digital Health Technology to Prevent and Treat Diabetes.

Using Digital Health Technology to Prevent and Treat Diabetes.

Parents’ ontological beliefs regarding the use of conversational agents at home: resisting the neoliberal discourse

ABSTRACT This paper develops a critical perspective on the use of conversational agents (CAs) with children at home. Drawing on interviews with eleven parents of pre-school children living in Norway, we illustrate the ways in which parents resisted the values epitomised by CAs. We problematise CAs’ attributes in light of parents’ ontological perceptions of what it means to be human and outline how their attitudes correspond to Bourdieu’s [1998a. Acts of Resistance. New York: New Press] concept of acts of resistance. For example, parents saw artificial conversation designed for profit as a potential threat to users’ autonomy and the instant gratification of CAs as a threat to children’s development. Parents’ antecedent beliefs map onto the ontological tensions between human and non-human attributes and challenge the neoliberal discourse by demanding freedom and equality for users rather than productivity and economic gain. Parents’ comments reflect the belief that artificial conversation with a machine inappropriately and ineffectively mimics a nuanced and intimate human-to-human experience in service of profit motives.

A conversational agent for emotion expression stimulation in persons with neurodevelopmental disorders

Difficulty in emotion expression and recognition is typical of the personality trait known as alexithymia, which is often observed in persons with neurodevelopmental disorders (NDD). Past research has investigated various forms of conversational technology for people with NDD, but only a few studies have explored the use of conversational agents to reduce alexithymia. This paper presents Emoty, a speech-based conversational agent designed for people with NDD to train emotional communication skills. An original characteristic of this agent is that it exploits the emotional expression power of the voice. Emoty engages users in small conversations during which they are asked to repeat sentences and express specific emotions using the appropriate vocal tone. We ran an empirical study to evaluate the usability and effectiveness of our conversational agent. The study involved 19 Italian individuals with NDD and alexithymia aged from 29 to 45 (10 women and 9 men). They used Emoty in five individual sessions over two and a half months. The results showed that two subjects encountered problems using the system because they had difficulty verbalizing the sentences and were not understood by Emoty. The others performed the assigned tasks with the agent. Their capability to express emotions with the voice consistently improved, and other benefits were observed in other social and communication skills.

Alexa as an Active Listener: How Backchanneling Can Elicit Self-Disclosure and Promote User Experience

Active listening is a well-known skill applied in human communication to build intimacy and elicit self-disclosure to support a wide variety of cooperative tasks. When applied to conversational UIs, active listening from machines can also elicit greater self-disclosure by signaling to the users that they are being heard, which can have positive outcomes. However, it takes considerable engineering effort and training to embed active listening skills in machines at scale, given the need to personalize active-listening cues to individual users and their specific utterances. A more generic solution is needed given the increasing use of conversational agents, especially by the growing number of socially isolated individuals. With this in mind, we developed an Amazon Alexa skill that provides privacy-preserving and pseudo-random backchanneling to indicate active listening. User study (N = 40) data show that backchanneling improves perceived degree of active listening by smart speakers. It also results in more emotional disclosure, with participants using more positive words. Perception of smart speakers as active listeners is positively associated with perceived emotional support. Interview data corroborate the feasibility of using smart speakers to provide emotional support. These findings have important implications for smart speaker interaction design in several domains of cooperative work and social computing.

The Development and Use of Chatbots in Public Health: Scoping Review.

BackgroundChatbots are computer programs that present a conversation-like interface through which people can access information and services. The COVID-19 pandemic has driven a substantial increase in the use of chatbots to support and complement traditional health care systems. However, despite the uptake in their use, evidence to support the development and deployment of chatbots in public health remains limited. Recent reviews have focused on the use of chatbots during the COVID-19 pandemic and the use of conversational agents in health care more generally. This paper complements this research and addresses a gap in the literature by assessing the breadth and scope of research evidence for the use of chatbots across the domain of public health.ObjectiveThis scoping review had 3 main objectives: (1) to identify the application domains in public health in which there is the most evidence for the development and use of chatbots; (2) to identify the types of chatbots that are being deployed in these domains; and (3) to ascertain the methods and methodologies by which chatbots are being evaluated in public health applications. This paper explored the implications for future research on the development and deployment of chatbots in public health in light of the analysis of the evidence for their use.MethodsFollowing the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines for scoping reviews, relevant studies were identified through searches conducted in the MEDLINE, PubMed, Scopus, Cochrane Central Register of Controlled Trials, IEEE Xplore, ACM Digital Library, and Open Grey databases from mid-June to August 2021. Studies were included if they used or evaluated chatbots for the purpose of prevention or intervention and for which the evidence showed a demonstrable health impact.ResultsOf the 1506 studies identified, 32 were included in the review. The results show a substantial increase in the interest of chatbots in the past few years, shortly before the pandemic. Half (16/32, 50%) of the research evaluated chatbots applied to mental health or COVID-19. The studies suggest promise in the application of chatbots, especially to easily automated and repetitive tasks, but overall, the evidence for the efficacy of chatbots for prevention and intervention across all domains is limited at present.ConclusionsMore research is needed to fully understand the effectiveness of using chatbots in public health. Concerns with the clinical, legal, and ethical aspects of the use of chatbots for health care are well founded given the speed with which they have been adopted in practice. Future research on their use should address these concerns through the development of expertise and best practices specific to public health, including a greater focus on user experience.

How humans obtain information from AI: Categorizing user messages in human-AI collaborative conversations

Although there is an increasingly number of research about the design and use of conversational agents, it is still difficult for conversational agents to completely replace human service. Therefore, more and more companies have adopted human-AI collaborative systems to deliver customer service. It is important to understand how people obtain information from human-AI collaborative conversations. While the existing work relies on self-reported methods to elicit qualitative feedback from users, we have concluded a categorization system for user messages in human-AI collaborative conversations after a thorough examination of a real-world customer service log, which could objectively reflect the user's information needs. We categorize user messages into five categories and 15 specific types related to three high-level intentions. Two annotators independently classified the same set of 1,478 user messages from 300 conversations and reached a moderate consistency. We summarize and report the characteristics of different message types and compare their usage in sessions with only human, AI, or both representatives. Our results show that different message types vary significantly in usage frequency, length, and text similarities with other messages in a session. Also, the frequency of using different message types in our dataset seems consistent over sessions with different types of representatives. But we also observed some significant differences in a few specific message types across the sessions with different representatives. Our results are used to suggest some areas for improvement and future work in human-AI collaborative conversational systems.

Designing Depression Screening Chatbots.

Advances in voice recognition, natural language processing, and artificial intelligence have led to the increasing availability and use of conversational agents (chatbots) in different settings. Chatbots are systems that mimic human dialogue interaction through text or voice. This paper describes a series of design considerations for integrating chatbots interfaces with health services. The present paper is part of ongoing work that explores the overall implementation of chatbots in the healthcare context. The findings have been created using a research through design process, combining (1) literature survey of existing body of knowledge on designing chatbots, (2) analysis on state-of-the-practice in using chatbots as service interfaces, and (3) generative process of designing a chatbot interface for depression screening. In this paper we describe considerations that would be useful for the design of a chatbot for a healthcare context.

Public Perceptions of Diabetes, Healthy Living, and Conversational Agents in Singapore: Needs Assessment.

BackgroundThe incidence of chronic diseases such as type 2 diabetes is increasing in countries worldwide, including Singapore. Health professional–delivered healthy lifestyle interventions have been shown to prevent type 2 diabetes. However, ongoing personalized guidance from health professionals is not feasible or affordable at the population level. Novel digital interventions delivered using mobile technology, such as conversational agents, are a potential alternative for the delivery of healthy lifestyle change behavioral interventions to the public.ObjectiveWe explored perceptions and experiences of Singaporeans on healthy living, diabetes, and mobile health (mHealth) interventions (apps and conversational agents). This study was conducted to help inform the design and development of a conversational agent focusing on healthy lifestyle changes.MethodsThis qualitative study was conducted in August and September 2019. A total of 20 participants were recruited from relevant healthy living Facebook pages and groups. Semistructured interviews were conducted in person or over the telephone using an interview guide. Interviews were transcribed and analyzed in parallel by 2 researchers using Burnard’s method, a structured approach for thematic content analysis.ResultsThe collected data were organized into 4 main themes: use of conversational agents, ubiquity of smartphone apps, understanding of diabetes, and barriers and facilitators to a healthy living in Singapore. Most participants used health-related mobile apps as well as conversational agents unrelated to health care. They provided diverse suggestions for future conversational agent-delivered interventions. Participants also highlighted several knowledge gaps in relation to diabetes and healthy living. Regarding barriers to healthy living, participants mentioned frequent dining out, high stress levels, lack of work-life balance, and lack of free time to engage in physical activity. In contrast, discipline, preplanning, and sticking to a routine were important for enabling a healthy lifestyle.ConclusionsParticipants in this study commonly used mHealth interventions and provided important insights into their knowledge gaps and needs in relation to changes in healthy lifestyle behaviors. Future digital interventions such as conversational agents focusing on healthy lifestyle and diabetes prevention should aim to address the barriers highlighted in our study and motivate individuals to adopt healthy lifestyle behavior.

MOCA: A Motivational Online Conversational Agent for Improving Student Engagement in Collaborative Learning

The use of conversational agents in computer-supported collaborative learning (CSCL) has been identified as a useful tactic for motivational intervention. The purpose of the current study was to design and implement a conversational agent called a motivational online conversational agent (MOCA) that incorporated motivational interviewing (MI) and was based on an intelligent dialog engine to enhance learner engagement in CSCL. Additionally, the study empirically examined the effects of MOCA on promoting positive changes in collaborative learning engagement through multiturn conversation interventions. A prototype system was developed by combining MOCA and an immersive virtual world, and an effectiveness study was conducted with 40 volunteers. A series of multilevel growth models based on the framework of the hierarchical linear model was established through multiwave longitudinal data. The results indicated that the use of MOCA significantly improved student engagement scores (<i>p</i> &lt; 0.001) and that female students performed better on collaborative tasks than male students (<i>p</i> &lt; 0.05, <i>t</i> &#x003D; 2.97). Additionally, time was an important predictor and significantly interacted with the MOCA-use condition. The study has implications for the design and assessment of conversational agents embodied in virtual reality.