Virtual Conversational Agents Research Articles

VCAs as partners or servants? The effects of information sensitivity and anthropomorphism roles on privacy concerns

Advances in machine learning and natural language processing have driven the growing popularity of virtual conversational agents (VCAs). This anthropomorphic communication approach relies on user information sharing and real-time feedback from VCAs, and has raised privacy concerns while affecting various social interactions and relationships. Previous research on reducing user privacy concerns has mainly focused on user information mining, sensitive user information requests and privacy policies, while little is known about the anthropomorphic roles of partners and servants at the human-machine social hierarchy level. Therefore, this study, based on human-computer interaction (service) anthropomorphism at social level, develops a framework to investigate the impact of information sensitivity and VCAs' anthropomorphic roles, including partner and servant, on users' privacy concerns, as well as the mediating effects of competence- and integrity-based trust. The results show that when highly sensitive information is requested, user privacy concerns are greater for a partner VCA than a servant VCA, and vice-versa. Meanwhile, when a VCA requests highly sensitive information, integrity-based trust mediates the relationship between servant VCAs and privacy concerns, and when a VCA requests low-sensitivity information, competence-based trust mediates the same relationship. These insights provide actionable implications for managers.

Older adults' experience with virtual conversational agents for health data collection.

Virtual conversational agents (i.e., chatbots) are an intuitive form of data collection. Understanding older adults' experiences with chatbots could help identify their usability needs. This quality improvement study evaluated older adults' experiences with a chatbot for health data collection. A secondary goal was to understand how perceptions differed based on length of chatbot forms. After a demographic survey, participants (≥60 years) completed either a short (21 questions), moderate (30 questions), or long (66 questions) chatbot form. Perceived ease-of-use, usefulness, usability, likelihood to recommend, and cognitive load were measured post-test. Qualitative and quantitative analyses were used. A total of 260 participants reported on usability and satisfaction metrics including perceived ease-of-use (5.8/7), usefulness (4.7/7), usability (5.4/7), and likelihood to recommend (Net Promoter Score = 0). Cognitive load (12.3/100) was low. There was a statistically significant difference in perceived usefulness between groups, with a significantly higher mean perceived usefulness for Group 1 than Group 3. No other group differences were observed. The chatbot was perceived as quick, easy, and pleasant with concerns about technical issues, privacy, and security. Participants provided suggestions to enhance progress tracking, edit responses, improve readability, and have options to ask questions. Older adults found the chatbot to be easy, useful, and usable. The chatbot required low cognitive load demonstrating it could be an enjoyable health data collection tool for older adults. These results will inform the development of a health data collection chatbot technology.

Chatbots in Online Marketing: A work in progressNajlae Zhani

In the digital era, chatbots have emerged as a transformative tool in online marketing (Crolic et al. 2022). This abstract introduces an ongoing research endeavor that specifically explores the role of chatbots in online marketing strategies. The research is a work in progress, aimed at understanding and contributing to the dynamic field of chatbot marketing. In today's dynamic digital landscape, chatbots have gained prominence as a powerful element of online marketing (Nwachukwu and Affen, 2023). This abstract serves as a preliminary exploration of the role of chatbots and conveys that our research is in progress, intended to provide insights into their evolving significance in online marketing. Chatbots, AI-driven virtual conversational agents, are becoming ubiquitous in digital marketing. Our research in progress focuses on understanding how these chatbots engage with customers, answer queries, and facilitate personalized interactions. We explore how chatbots enhance customer experience and drive engagement on websites and messaging platforms. A key aspect of our research in progress is dedicated to conversational marketing, wherein chatbots play a pivotal role in guiding customers through their journey. We investigate how chatbots create a seamless conversational experience and the impact on lead generation and conversions. Our research explores how chatbots collect and analyze data to offer personalized recommendations and insights. We aim to understand how chatbots leverage data-driven personalization to enhance user experience and provide tailored solutions to customers. As part of our research in progress, we delve into the future of chatbots in marketing. We anticipate discussing emerging trends and potential challenges, such as the integration of chatbots with emerging technologies, including voice assistants and augmented reality. Methodology and Data Collection Our ongoing research employs a mixed-method approach, combining quantitative and qualitative data collection techniques. We conduct surveys, interviews, and user behavior analysis to gather insights into how chatbots are perceived and utilized by businesses and consumers. Implications for Businesses Our research aims to offer practical insights for businesses aiming to leverage chatbots in their marketing strategies. We discuss potential benefits and challenges, providing guidance on the optimal use of chatbots to drive brand engagement and customer satisfaction. In conclusion, this abstract represents an ongoing research initiative focusing on the pivotal role of chatbots in online marketing. By emphasizing that our research is a work in progress, we signal our commitment to contributing to the evolving field of chatbot marketing. As we continue our research journey, we anticipate providing valuable insights to assist businesses in harnessing the potential of chatbots for online marketing success.

Development of Chatbot-Psychologist: Dataset, Architecture, Design and Chatbot in Use

Virtual assistants are software applications that allow users to have conversations with the software in the same way as they would in real life. Creating functional conversational agents have been the most challenging aspect of working with AI ever since the field was first introduced. Virtual conversational agents' primary responsibility is to properly interpret natural communication and reply in an acceptable manner, in spite of the fact that they may do a range of activities. In the past, AI agents were created either via the use of handwritten rules and commands or by straightforward statistical approaches. End-to-end Automation has mostly supplanted these models because of its superior potential for learning new skills. The encoder-decoder deep learning approach is currently the method that has the greatest amount of interest for simulating conversations. The domain of linguistic understanding served as a source of motivation for the development of this concept. In this article, we present an overview o the findings from our study into the development of an immersive digital conversational agent that might potentially provide sufferers psychological aid. We used Rasa NLU approach, which is based on NLP practices, in order to construct and train the chatbot. The findings of the study indicated that providing appropriate replies to patients' questions and concerns during conversations had a prediction accuracy about 75 percent.

Virtual conversational agents versus online forms: Patient experience and preferences for health data collection.

ObjectiveVirtual conversational agents, or chatbots, have emerged as a novel approach to health data collection. However, research on patient perceptions of chatbots in comparison to traditional online forms is sparse. This study aimed to compare and assess the experience of completing a health assessment using a chatbot vs. an online form.MethodsA counterbalanced, within-subject experimental design was used with participants recruited via Amazon Mechanical Turk (mTurk). Participants completed a standardized health assessment using a chatbot (i.e., Dokbot) and an online form (i.e., REDCap), each followed by usability and experience questionnaires. To address poor data quality and preserve integrity of mTurk responses, we employed a thorough data cleaning process informed by previous literature. Quantitative (descriptive and inferential statistics) and qualitative (thematic analysis and complex coding query) approaches were used for analysis.ResultsA total of 391 participants were recruited, 185 of whom were excluded, resulting in a final sample size of 206 individuals. Most participants (69.9%) preferred the chatbot over the online form. Average Net Promoter Score was higher for the chatbot (NPS = 24) than the online form (NPS = 13) at a statistically significant level. System Usability Scale scores were also higher for the chatbot (i.e. 69.7 vs. 67.7), but this difference was not statistically significant. The chatbot took longer to complete but was perceived as conversational, interactive, and intuitive. The online form received favorable comments for its familiar survey-like interface.ConclusionOur findings demonstrate that a chatbot provided superior engagement, intuitiveness, and interactivity despite increased completion time compared to online forms. Knowledge of patient preferences and barriers will inform future design and development of recommendations and best practice for chatbots for healthcare data collection.

The influence of religiosity on consumers' evaluations of brands using artificial intelligence

AbstractPrior research is limited that examines how consumers respond to artificial intelligence (AI) used by businesses and in marketing. Specifically, more research is needed that explores which consumers are more likely to support AI use in marketing as well as why this occurs to develop targeted marketing campaigns toward these consumers. Through three studies, our research addresses these gaps by testing the role of religiosity on AI evaluations. Study 1 shows that religiosity positively influences perceptions of AI in marketing, which is mediated by greater trust in AI. Study 2 primes religiosity to demonstrate the causality of the effects, showing that consumers exposed to a religiosity prime more positively evaluate companies and their employees when AI is used. Study 3 then better confirms the underlying mechanisms of these effects using a virtual conversational agent context by showing that higher religiosity consumers are more trusting of the unseen, which leads to greater trust of a company, and ultimately positively influences evaluations of companies using AI. We contribute to belief congruence theory in identifying the role of trust (specifically trust in the unseen related to AI) in the belief‐behavior relationship. We also identify how marketers can use religious cues to increase trust in AI in marketing.

An empirical study investigating the user acceptance of a virtual conversational agent interface for family health history collection among the geriatric population.

Critical for the early diagnosis of genetic disorders, a Family Health History (FHx) can be collected in several ways including electronic FHx tools, which aid easy editing and sharing by linking with other information management portals. The user acceptance of such systems is critical, especially among older adults experiencing motor and cognitive issues. This study investigated two types of FHx interfaces, standard and Virtual Conversational Agent (VCA), using 30 young (between 18 and 30) and 24 older participants (over 60). Workload, usability and performance data were collected. Even though participants required less time to complete three of five tasks on the standard interface, the VCA interface performed better in terms of subjective workload and usability. Additionally, 67% of the older adults preferred the VCA interface since it provided context-based guidance during the data collection process. The results from this study have implications for the use of virtual assistants in FHx and other areas of data collection.

Privacy Enhanced Multimodal Neural Representations for Emotion Recognition

Many mobile applications and virtual conversational agents now aim to recognize and adapt to emotions. To enable this, data are transmitted from users' devices and stored on central servers. Yet, these data contain sensitive information that could be used by mobile applications without user's consent or, maliciously, by an eavesdropping adversary. In this work, we show how multimodal representations trained for a primary task, here emotion recognition, can unintentionally leak demographic information, which could override a selected opt-out option by the user. We analyze how this leakage differs in representations obtained from textual, acoustic, and multimodal data. We use an adversarial learning paradigm to unlearn the private information present in a representation and investigate the effect of varying the strength of the adversarial component on the primary task and on the privacy metric, defined here as the inability of an attacker to predict specific demographic information. We evaluate this paradigm on multiple datasets and show that we can improve the privacy metric while not significantly impacting the performance on the primary task. To the best of our knowledge, this is the first work to analyze how the privacy metric differs across modalities and how multiple privacy concerns can be tackled while still maintaining performance on emotion recognition.

Interpreting Multimodal Machine Learning Models Trained for Emotion Recognition to Address Robustness and Privacy Concerns

Many mobile applications and virtual conversational agents now aim to recognize and adapt to emotions. These predicted emotions are used in variety of downstream applications: (a) generating more human like dialogues, (b) predicting mental health issues, and (c) hate speech detection and intervention. To enable this, data are transmitted from users' devices and stored on central servers. These data are then processed further, either annotated or used as inputs for training a model for a specific task. Yet, these data contain sensitive information that could be used by mobile applications without user's consent or, maliciously, by an eavesdropping adversary. My work focuses on two major issues that are faced while training emotion recognition algorithms: (a) privacy of the generated representations and, (b) explaining and ensuring that the predictions are robust to various situations. Tackling these issues would lead to emotion based algorithms that are deployable and helpful at a larger scale, thus enabling more human like experience when interacting with AI.

Family health history collected by virtual conversational agents: An empirical study to investigate the efficacy of this approach.

Family health history (FHx) is one of the simplest and most cost-effective and efficient ways to collect health information that could help diagnose and treat genetic diseases at an early stage. This study evaluated the efficacy of collecting such family health histories through a virtual conversational agent (VCA) interface, a new method for collecting this information. Standard and VCA interfaces for FHx collection were investigated with 50 participants, recruited via email and word of mouth, using a within-subject experimental design with the order of the interfaces randomized and counterbalanced. Interface workload, usability, preference, and satisfaction were assessed using the NASA Task Load Index workload instrument, the IBM Computer System Usability Questionnaire, and a brief questionnaire derived from the Technology Acceptance Model. The researchers also recorded the number of errors and the total task completion time. It was found that the completion times for 2 of the 5 tasks were shorter for the VCA interface than for the standard one, but the overall completion time was longer (17min 44s vs. 16min 51s, p=.019). We also found the overall workload to be significantly lower (34.32 vs. 42.64, p=.003) for the VCA interface, and usability metrics including overall satisfaction (5.62 vs. 4.72, p<.001), system usefulness (5.76 vs. 4.84, p=.001), information quality (5.43 vs. 4.62, p<.001), and interface quality (5.66 vs. 4.64, p<.001) to be significantly higher for this interface as well. Approximately 3 out of 4 participants preferred the VCA interface to the standard one. Although the overall time taken was slightly longer than with standard interface, the VCA interface was rated significantly better across all other measures and was preferred by the participants. These findings demonstrate the advantages of an innovative VCA interface for collecting FHx, validating the efficacy of using VCAs to collect complex patient-specific data in health care.

Construction of a Virtual Ballroom Dance Instructor

The use of virtual conversational agents is awaited in the tutoring of physical skills such as sports or dances. This paper describes an ongoing project aiming to realize a virtual instructor for ballroom dance. First, a human-human experiment is conducted to collect the interaction corpus between a professional instructor and six learners. The verbal and non-verbal behaviors of the instructor is analyzed and served as the base of a state transition model for ballroom dance tutoring. In order to achieve natural and highly interactive instruction during the multi-modal interaction between the virtual instructor and the learner, it is necessary to divide the learner's motion into small but meaningful segments in real-time. In the case of ballroom dance, the smallest meaningful unit of dance steps is called count which should be synchronized with the beats of accompanied music. A method of automatic extraction from the learner's dance practice motion into count segments is proposed in this paper. This method is based on trajectory similarity comparison between the motion of the learner and the data collected from a professional dance instructor using Angular Matrix for Shape Similarity (AMSS). Another algorithm for identifying the worst performed portion and the corresponding improvement of the learner's motion is also proposed. Finally, a comparison experiment between the prototype system and a non-interactive self-training system was conducted and reported.

The potential benefits of multi_modal social interaction on the web for senior users

The Internet has become a key factor in today's society for improving lifestyles, from entertainment to obtaining every day life necessities. Despite this, most seniors hesitate to use the Internet. As a result, there is a lower growth rate in Internet usage among seniors compared to those of younger populations. In this research, a web interface using a virtual conversational agent has been designed to address the problems of existing web interfaces; especially e-commerce sites. Data from a comparative study between an avatar design and a traditional design for a web-based retail store has been collected and analyzed. The results support the avatar design over a traditional web design; however, in some cases, the evidence collected showed no statistical significance.

Gaze, conversational agents and face-to-face communication

In this paper, we describe two series of experiments that examine audiovisual face-to-face interaction between naive human viewers and either a human interlocutor or a virtual conversational agent. The main objective is to analyze the interplay between speech activity and mutual gaze patterns during mediated face-to-face interactions. We first quantify the impact of deictic gaze patterns of our agent. We further aim at refining our experimental knowledge on mutual gaze patterns during human face-to-face interaction by using new technological devices such as non-invasive eye trackers and pinhole cameras, and at quantifying the impact of a selection of cognitive states and communicative functions on recorded gaze patterns.