Novel role-wise attention mechanism for predicting purchases made through chat-based online customer services

2023 | Decision Support Systems | Citations: 0

Authors: Wang, Yajing; Chen, Gang; Xu, Yunjie (Calvin); Lu, Xianghua

Abstract: Online chat has become a key channel for firms to provide pre-purchase customer ... Expand

Abstract: Online chat has become a key channel for firms to provide pre-purchase customer services and generate pur­ chases. Firms must deploy their service personnel based on the customers' purchase probability. Conversational contingency is indicative of purchase probability. However, the unstructured nature of textual data hinders the use of interaction information. We develop a novel role-wise attention mechanism to capture the conversational contingency among customers and service agents. Our results reveal that the role-wise attention mechanism is effective for measuring service interaction quality. When the mechanism is combined with a deep-learning model, it outperforms traditional self-attention mechanisms for predicting customer purchases. The area under the curve of the model is 0.85. Ablation studies indicate that external attention, a core element of the role-wise attention mechanism, substantially improves prediction performance by capturing the conversational contin­ gency between communicators. Collapse

Topics: customer service social media online chat Python service-support information system

Methods: machine learning BERT deep learning design artifact ablation analysis

Anthropomorphism of AI-based Intelligent Customer Service, and Its Affective and Behavioral Consequences

2022 | HICSS | Citations: 0

Authors: Chang-Chien, Yu-Ting; Cheng, Jacob Chun; Hsu, Shih-Chieh; Yeh, Yi Wen

Abstract: Recently, as many users turn to social media to interact with service providers, ... Expand

Abstract: Recently, as many users turn to social media to interact with service providers, organizations apply Artificial intelligence (AI) to improve the efficiency and effectiveness of the operation. This type of customer service system is called intelligent customer service (ICS) which one of the most commonly adopted tools is chatbot. Since chatbot is AI-empowered, whether this system can effectively interact with customers and solve their problems is critical. However, the quality of ICS has received significant attention recently, and a lack of systematic study on the outcomes of anthropomorphism leaves this question unanswered in an ICS context. Based on a cognitive-affective-behavioral framework, this study attempts to understand whether anthropomorphism can promote desired behaviors (including usage and citizenship behaviors) through enhancing affective outcomes, such as satisfaction and identity. Data collected from 183 chatbotICS users, this study illustrates how anthropomorphism can increase quality, enhance satisfaction and identity. Furthermore, we also show that satisfaction and identity lead to further usage and citizenship behaviors. This highlights the importance of increasing anthropomorphism for the chatbot-ICS. Collapse

Topics: customer service artificial intelligence service-support information system robotic system quality

Methods: chatbot survey factor analysis descriptive statistic cross sectional research

Racial Bias in Customer Service: Evidence from Twitter

2022 | Information Systems Research | Citations: 6

Authors: Gunarathne, Priyanga; Rui, Huaxia; Seidmann, Abraham

Abstract: This paper provides the first large-scale evidence of business-to-customer racia ... Expand

Abstract: This paper provides the first large-scale evidence of business-to-customer racial bias (B2C bias) on a digital platform, on which the perpetrators are individual employees who act on behalf of a company and the victims are customers. This is in contrast to existing studies of racial bias on digital platforms that focus on peer-to-peer marketplaces (e.g., eBay), in which both the perpetrators and the victims are individuals acting independently and on their own behalf. In particular, we present the first evidence of B2C bias in corporate social media customer service, a practice that has grown in popularity recently. Unlike traditional call centers, agents providing customer service on social media respond, on average, to fewer than half of the complaints they receive as per our analysis. We investigate the effect of a complaining customer’s racial identity, as revealed by the social media profile picture, on the chance of receiving a response. By analyzing more than 57,000 social media customer complaints to major U.S. airlines and leveraging a variety of analytics techniques, including text mining and facial recognition, we present quantitative evidence that African American customers are less likely to receive a response when they complain than otherwise similar White customers. Furthermore, our deep learning–based falsification test shows that the bias is absent without the visual cue that reveals racial identity. This study offers a practical yet powerful recommendation for companies: conceal all customer profile pictures from their employees while delivering social media customer service. Collapse

Topics: Twitter social media customer service application programming interface business-to-consumer

Methods: experimental group deep learning biometric measurement facial recognition system econometric modeling

Can we Help the Bots? Towards an Evaluation of their Performance and the Creation of Human Enhanced Artifact for Emotions De-escalation

2022 | HICSS | Citations: 0

Authors: Palese, Biagio; Pickard, Matthew; Bartosiak, Marcin

Abstract: We propose a hybrid intelligence socio-technical artifact that identifies a thre ... Expand

Abstract: We propose a hybrid intelligence socio-technical artifact that identifies a threshold where the chatbot requires human intervention in order to continue to perform at an appropriate level to achieve the predefined objective of the system. We leverage the Yield Shift Theory of Satisfaction, the Intervention Theory and the Nudge Theory to develop meta requirements and design principles for this system. We discuss the first iteration of implementation and evaluation of the artifact components. Collapse

Topics: application programming interface system support digital dashboard crowdsourcing gamification

Methods: chatbot design artifact random forest classification kernel theory design requirement

Theories: yield shift theory of satisfaction nudge theory

Integration of AI into Customer Service: A Taxonomy to Inform Design Decisions

2022 | European Conference On Information Systems | Citations: 4

Authors: Poser, Mathis; Wiethof, Christina; Bittner, Eva

Abstract: Artificial Intelligence (AI) is increasingly deployed in customer service for v ... Expand

Abstract: Artificial Intelligence (AI) is increasingly deployed in customer service for various service delivery tasks. Research and practice alike have extensively dealt with the use, benefits, and effects of AI solutions in customer service contexts. Nevertheless, knowledge on AI integration is dispersed and unsystematized. This paper addresses this gap by presenting a taxonomy to inform design decisions for the integration of AI into customer service with five meta-dimensions, 12 dimensions, and 32 characteristics. Through a rigorous and systematic development process comprising multiple iterations and evaluation episodes, state-of-the-art AI solutions from practice and the current state of knowledge from research were systematized to classify AI use cases. Thus, we contribute with systemized design knowledge to, both, the theoretical knowledge base as well as to practice for application. Eventually, we disclose future research avenues addressing certain meta-dimensions as well as the extension of the taxonomy itself. Collapse

Topics: artificial intelligence customer service database system service-support information system knowledge base

Methods: personal interview survey design science literature study structured literature research

Service Failure Recovery in the Sharing Economy: A Case Study of Airbnb

2021 | HICSS | Citations: 0

Authors: Sosa, Luis; Abhari, Kaveh; Pesavento, Michael; Koobchehr, Farzan; Xiao, Bo; Kohsuwan, Pocky; Olivares, Fernando

Abstract: Sharing platforms are becoming increasingly common, revolutionizing how peers in ... Expand

Abstract: Sharing platforms are becoming increasingly common, revolutionizing how peers interact and share resources across an array of online applications. While the sharing economy itself is established, less is known about service failures and corresponding recovery strategies that are relevant to it. This research investigates the myriad effects of service failures (and their associated recovery strategies) on customer experience in the digital sharing economy. Findings suggest that different service failure strategies exert differing effects on customer experience, which subsequently affects the behavior towards the service being provided and the service provider. The suggestions given here respond to important implications for research and practitioners by offering new ways to explore and detect service failures and possible recovery strategies. Collapse

Topics: Airbnb sharing platform digital platform customer service IS failure

Methods: qualitative coding case study qualitative content analysis longitudinal research

“What can I(S) do for you?”: How technology enables service providers to elicit customers’ preferences and deliver personalized service

2020 | Information & Management | Citations: 0

Authors: Bonaretti, Dario; Bartosiak, Marcin; Lui, Tsz-Wai; Piccoli, Gabriele; Marchesani, Daniele

Abstract: This research studies the effective use of an IT-enabled Customer Service System ... Expand

Abstract: This research studies the effective use of an IT-enabled Customer Service System (CSS) in the hospitality sector. Drawing on Effective Use Theory, we study how digital representations of customers’ preferences support service providers in the service personalization process. Our model posits customers’ preferences elicitation as the antecedent of CSS’ effective use, which yields superior customer value. This representational perspective matches recent efforts to study customer value as cocreated by customers and service providers. Based on a longitudinal dataset of 118,647 hotel reservations, we explain how digital representations of customers’ preferences enable service providers to personalize service and increase customer value. Collapse

Topics: usability personalization information system use participatory design online review

Methods: design methodology longitudinal research chi squared test logistic regression archival research

Theories: representation theory service-dominant logic

A Platform for Value Co-creation in SME Networks

2020 | DESRIST | Citations: 0

Authors: Hönigsberg, Sarah

Abstract: The design of digital value co-creation (VCC) platforms is an increasingly promi ... Expand

Abstract: The design of digital value co-creation (VCC) platforms is an increasingly prominent research topic. The creation and introduction of these platforms are transforming the cooperation of entire company networks and ecosystems and is a challenging endeavor. In this paper, a multi-year action design research (ADR) study in a small and medium-sized enterprise (SME) network in the German textile industry is presented. As with many SME networks, the creation of value and innovation is hindered by discontinuous digitalization. This leads to major problems in the knowledge-intensive and complex development of technical textiles. A digital VCC platform mitigates production delays, interruptions, and errors under the competitive pressure of the textile industry, which is critical for the survival of such networks. For a lack of suitable solutions, a VCC platform is cyclically designed and evaluated in our network. Design principles and features are derived to address resource integration challenges in service innovation in SME networks. Collapse

Topics: business process innovation innovation management digital platform knowledge base value creation

Methods: action design research design principle design science design theory design methodology

Theories: service-dominant logic

Business Practice of Social Media - Platform and Customer Service Adoption

2020 | International Conference on Information Systems | Citations: 0

Authors: Sun, Shujing; Gao, Yang; Rui, Huaxia

Abstract: This paper examines the key drivers in business adoptions of the platform and cu ... Expand

Abstract: This paper examines the key drivers in business adoptions of the platform and customer service within the context of social media. We carry out the empirical analyses using the decision trajectories of the international airline industry on Twitter. We find that a firm’s decision-making is subject to both peer influence and consumer pressure. Regarding peer influence, we find that the odds of both adoptions increase when the extent of peers’ adoption increases. We also identify the distinctive role of consumers. Specifically, before the platform adoption, firms learn about potential consequences from consumer reactions to peers’ adoptions. Upon the platform adoption, consumer voices directed at a firm itself is more crucial to customer service adoption. Furthermore, while both positive and neutral voices facilitate platform adoption, only positive voices significantly contribute to customer service adoption. The findings confirm the distinct trade-offs faced by firms at different adoption stages: while firms are motivated to adopt the platform to reach potential customers, firms care more about the online reputation when deciding customer service adoption. Collapse

Topics: Twitter social media customer service decision making marketing management

Methods: survival analysis hazard function longitudinal research

Theories: theory of everything institutional theory

Heuristic Theorizing in Software Development: Deriving Design Principles for Smart Glasses-based Systems

2019 | International Conference on Business Informatics | Citations: 0

Authors: Berkemeier, Lisa; Zobel, Benedikt; Werning, Sebastian; Vogel, Jannis; Ickerott, Ingmar; Thomas, Oliver

Abstract: Design knowledge on smart glasses-based systems is scarce. Utilizing literature ... Expand

Abstract: Design knowledge on smart glasses-based systems is scarce. Utilizing literature analysis on software development publications, insights from the design and implementation of four smart glasses-based systems and expert interviews, we elicited 16 design principles to provide guidance in the development of future service support systems. Heuristic Theorizing is an abductive Design Science Research method, hitherto far too little known or little noticed, which was applied to conduct the research. We contribute to theory and practice with applicable design principles to support the development of smart glasses-based systems. Phenomena known to have an impact on the adoption of smart glasses are addressed by these design principles. Collapse

Topics: smart glass logistics management usability systems development computer hardware

Methods: design principle design artifact qualitative interview theory development theoretical contribution