Artificial Intelligence-based Assistants
Permanent URI for this collection
Browse
Recent Submissions
Item Whose Advice Counts More – Man or Machine? An Experimental Investigation of AI-based Advice Utilization(2021-01-05) Mesbah, Neda; Tauchert, Christoph; Buxmann, PeterDue to advances in Artificial Intelligence (AI), it is possible to provide advisory services without human advisors. Derived from judge-advisor system literature, we examined differences in the advice utilization depending on whether it is given by an AI-based or human advisor and the similarity of the advice and their own estimation. Drawing on task-technology fit we investigated the relationship between task, advisor and advice utilization. In study A we measured the actual advice utilization within a guessing game and in study B we measured the perceived task-advisor fit for this game. The findings show that compared to human advisors, judges utilize advices of AI-based advisors more when the advice is similar to their own estimation. When the advice is very different to their estimation, the advices are used equally. Concluding, we investigated AI-based advice utilization and presented insights for professionals providing AI-based advisory services.Item The Role of Trusting Beliefs in Voice Assistants during Voice Shopping(2021-01-05) Mari, Alex; Algesheimer, RenéArtificial intelligence-based voice assistants (VAs) such as Amazon Alexa deliver personalized product recommendations in order to match consumers’ needs. The use of voice assistants for shopping purposes incorporates elements of risk affecting when and how they are considered trusted relationship partners. In this uncertain environment, it is unclear ‘when’ voice assistants are capable of gaining trust and ‘how’ the development of such a trusted relationship affects decisions. This research explores the effect of trusting beliefs towards voice assistants on decision satisfaction through the indirect effect of consideration set size (n. of options), in the context of voice shopping. Findings of an individual-session online experiment (N = 180) show a positive direct effect of trust on customer’s satisfaction and a mediating role of set size, confirming consumers’ bias towards default choices. This study highlights the consequences of trust in AI-enabled voice assistants for decision-making during utilitarian purchases.Item Practical Challenges of Virtual Assistants and Voice Interfaces in Industrial Applications(2021-01-05) Gärtler, Marco; Schmidt , BenediktVirtual assistant systems promise ubiquitous and simple access to information, applications and physical appliances. Their foundation on intent-oriented queries and support of natural language makes them an ideal tool for human-centric application. The general approach to build such systems as well as the main building blocks are well-understood and offered as off-the-shelf components. While there are prominent examples in the service sector, other sectors such as the manufacturing and process industries have nothing comparable. We investigate the practical challenges to build a virtual assistant using a representative and simplified case from the domain of knowledge retrieval. A qualitative study reveals two major obstacles: Firstly, a high level of expectations from users and, secondly, a disproportional amount of effort to get all details and having a robust system. Overall, implementing a virtual assistant for an industrial application is technical feasible, yet requires significant effort and understanding of the target audience.Item Machinelike or Humanlike? A Literature Review of Anthropomorphism in AI-Enabled Technology(2021-01-05) Li, Mengjun; Suh, AyoungDue to the recent proliferation of AI-enabled technology (AIET), the concept of anthropomorphism, human likeness in technology, has increasingly attracted researchers’ attention. Researchers have examined how anthropomorphism influences users’ perception, adoption, and continued use of AIET. However, researchers have yet to agree on how to conceptualize and operationalize anthropomorphism in AIET, which has resulted in inconsistent findings. A comprehensive understanding is thus needed of the current state of research on anthropomorphism in AIET contexts. To conduct an in-depth analysis of the literature on anthropomorphism, we reviewed 35 empirical studies focusing on conceptualizing and operationalizing AIET anthropomorphism, and its antecedents and consequences. Based on our analysis, we discuss potential research gaps and offer directions for future research.Item Communicating with Machines: Conversational Agents with Personality and the Role of Extraversion(2021-01-05) Ahmad, Rangina; Siemon, Dominik; Robra-Bissantz, SusanneCommunication with conversational agents (CA) has become increasingly important. It therefore is crucial to understand how individuals perceive interaction with CAs and how the personality of both the CA and the human can affect the interaction experience. As personality differences are manifested in language cues, we investigate whether different language style manifestations of extraversion lead to a more anthropomorphized perception (specifically perceived humanness and social presence) of the personality bots. We examine, whether individuals rate communication satisfaction of a CA similar to their own personality as higher (law of attraction). The results of our experiment indicate that highly extraverted CAs are generally better received in terms of social presence and communication satisfaction. Further, incorporating personality into CAs increases perceived humanness. Although no significant effects could be found in regard to the law of attraction, interesting findings about ambiverts could be made. The outcomes of the experiment contribute towards designing personality-adaptive CAs.Item All About the Name: Assigning Demographically Appropriate Names to Data-Driven Entities(2021-01-05) Jung, Soon-Gyo; Salminen, Joni; Jansen, Bernard J.We develop a method for assigning demographically appropriate names to data-driven entities, such as personas, chatbots, and virtual agents. The value of this method is removing the time-consuming human effort in this task. To demonstrate our method, we collect four million user profiles with gender, age, and country information from an international online social network. From this dataset, we obtain 1, 031, 667 unique names covering 3, 088 demographic group combinations that our method considers as gender, age, and nationality appropriate. A manual evaluation by raters from 34 countries shows a demographic appropriateness score of 85.6%. The demographically appropriate names can be utilized for data-driven personas, virtual agents, chatbots, and other humanized entities.Item A Conceptual Model for Assistant Platforms(2021-01-05) Schmidt, Rainer; Alt, Rainer; Zimmermann, AlfedAssistant platforms are becoming a key element for the business model of many companies. They have evolved from assistance systems that provide support when using information (or other) systems to platforms in their own. Alexa, Cortana or Siri may be used with literally thousands of services. From this background, this paper develops the notion of assistant platforms and elaborates a conceptual model that supports businesses in developing appropriate strategies. The model consists of three main building blocks, an architecture that depicts the components as well as the possible layers of an assistant platform, the mechanism that determines the value creation on assistant platforms, and the ecosystem with its network effects, which emerge from the multi-sided nature of assistant platforms. The model has been derived from a litera-ture review and is illustrated with examples of existing assistant platforms. Its main purpose is to advance the understanding of assistant platforms and to trigger future research.Item Introduction to the Minitrack on Artificial Intelligence-based Assistants(2021-01-05) Schmidt, Rainer; Alt, Rainer; Zimmermann, Alfed