Learning with Digital Agents: An Analysis based on the Activity Theory

2023 | Journal of Management Information Systems | Citations: 0

Authors: Dolata, Mateusz; Katsiuba, Dzmitry; Wellnhammer, Natalie; Schwabe, Gerhard

Abstract: Digital agents are considered a general-purpose technology. They spread quickly ... Expand

Abstract: Digital agents are considered a general-purpose technology. They spread quickly in private and organizational contexts, including education. Yet, research lacks a conceptual framing to describe interaction with such agents in a holistic manner. While focusing on the interaction with a pedagogical agent, that is, a digital agent capable of natural-language interaction with a learner, we propose a model of learning activity based on activity theory. We use this model and a review of prior research on digital agents in education to analyze how various characteristics of the activity, including features of a pedagogical agent or learner, influence learning outcomes. The analysis leads to identification of information systems research directions and guidance for developers of pedagogical agents and digital agents in general. We conclude by extending the activity theory-based model beyond the context of education and show how it helps designers and researchers ask the right questions when creating a digital agent. Collapse

Topics: online learning system artificial intelligence website human AI collaboration human computer interaction

Methods: literature study design artifact literature sample experiment structured literature research

Theories: activity theory socio technical theory theory of intentionality

Incorporating Multiple Knowledge Sources for Targeted Aspect-based Financial Sentiment Analysis

2023 | ACM Transactions on Management Information Systems | Citations: 0

Authors: Du, Kelvin; Xing, Frank; Cambria, Erik

Abstract: Combining symbolic and subsymbolic methods has become a promising strategy as re ... Expand

Abstract: Combining symbolic and subsymbolic methods has become a promising strategy as research tasks in AI grow increasingly complicated and require higher levels of understanding. Targeted Aspect-based Financial Sentiment Analysis (TABFSA) is an example of such complicated tasks, as it involves processes like information extraction, information specification, and domain adaptation. However, little is known about the design principles of such hybrid models leveraging external lexical knowledge. To fill this gap, we define anterior, parallel, and posterior knowledge integration and propose incorporating multiple lexical knowledge sources strategically into the fine-tuning process of pre-trained transformer models for TABFSA. Experiments on the Financial Opinion mining and Question Answering challenge (FiQA) Task 1 and SemEval 2017 Task 5 datasets show that the knowledge-enabled models systematically improve upon their plain deep learning counterparts, and some outperform state-of-the-art results reported in terms of aspect sentiment analysis error. We discover that parallel knowledge integration is the most effective and domain-specific lexical knowledge is more important according to our ablation analysis. Collapse

Topics: microblog domain-specific language Twitter social media knowledge graph

Methods: sentiment analysis RoBERTa transformer model BERT XLNet

Emotional Synchrony and Viewers’ Consumption: Evidence from Live Streaming of Virtual Idols

2023 | International Conference on Information Systems | Citations: 0

Authors: Duan, Menghan; Zhang, Yueyue; Zhang, Qi; Zhang, Cheng

Abstract: Live streaming has become increasingly popular in recent years for users to bro ... Expand

Abstract: Live streaming has become increasingly popular in recent years for users to broadcast live events or experiences to an online audience in real-time. With the prevalence of the metaverse, virtual idols appear on the stage of live streaming. Unlike real human streamers, virtual idols are computer-generated avatars whose appearance does not exist in the physical world. Though existing literature has documented the effect of emotional displays of human streamers on viewers' behavior, whether emotional displays of avatar streamers affect viewers' behavior, especially viewers' consumption behavior has not yet been explored. To fill this gap, we focus on the live streaming of virtual idols and conduct deep learning approaches to investigate the emotional synchrony between avatar streamers and their viewers based on moment-tomoment data. Our analysis of large-scale video data shows that the larger the discrepancy between the streamers' acoustic-and text-emotions, the more likely the viewers will pay. This study sheds light on how avatars can be designed to leverage emotion to engage viewers in more consumption. Collapse

Topics: marketing management human AI collaboration microblog open source Python

Methods: deep learning computational algorithm XGBoost machine learning gradient boosting

Opinion Paper: “So what if ChatGPT wrote it?” Multidisciplinary perspectives on opportunities, challenges and implications of generative conversational AI for research, practice and policy

2023 | International Journal of Information Management | Citations: 0

Authors: Dwivedi, Yogesh K.; Kshetri, Nir; Hughes, Laurie; Slade, Emma Louise; Jeyaraj, Anand; Kar, Arpan Kumar; Baabdullah, Abdullah M.; Koohang, Alex; Raghavan, Vishnupriya; Ahuja, Manju; Albanna, Hanaa; Albashrawi, Mousa Ahmad; Al-Busaidi, Adil S.; Balakrishnan, Janarthanan; Barlette, Yves; Basu, Sriparna; Bose, Indranil; Brooks, Laurence; Buhalis, Dimitrios; Carter, Lemuria; Chowdhury, Soumyadeb; Crick, Tom; Cunningham, Scott W.; Davies, Gareth H.; Davison, Robert M.; Dé, Rahul; Dennehy, Denis; Duan, Yanqing; Dubey, Rameshwar; Dwivedi, Rohita; Edwards, John S.; Flavián, Carlos; Gauld, Robin; Grover, Varun; Hu, Mei-Chih; Janssen, Marijn; Jones, Paul; Junglas, Iris; Khorana, Sangeeta; Kraus, Sascha; Larsen, Kai R.; Latreille, Paul; Laumer, Sven; Malik, F. Tegwen; Mardani, Abbas; Mariani, Marcello; Mithas, Sunil; Mogaji, Emmanuel; Nord, Jeretta Horn; O’Connor, Siobhan; Okumus, Fevzi; Pagani, Margherita; Pandey, Neeraj; Papagiannidis, Savvas; Pappas, Ilias O.; Pathak, Nishith; Pries-Heje, Jan; Raman, Ramakrishnan; Rana, Nripendra P.; Rehm, Sven-Volker; Ribeiro-Navarrete, Samuel; Richter, Alexander; Rowe, Frantz; Sarker, Suprateek; Stahl, Bernd Carsten; Tiwari, Manoj Kumar; Van Der Aalst, Wil; Venkatesh, Viswanath; Viglia, Giampaolo; Wade, Michael; Walton, Paul; Wirtz, Jochen; Wright, Ryan

Abstract: ☆ This editorial opinion paper provides a subjective viewpoint on the potential ... Expand

Abstract: ☆ This editorial opinion paper provides a subjective viewpoint on the potential impact of generative AI technologies such as ChatGPT in the domains of education, business, and society. Its objective is to offer initial guidance on the opportunities, challenges, and implications associated with these technologies. It is worth noting that, given its nature as an editorial opinion piece, this submission has not undergone a formal double-blind review process but has been reviewed informally by appropriate experts.Transformative artificially intelligent tools, such as ChatGPT, designed to generate sophisticated text indistinguishable from that produced by a human, are applicable across a wide range of contexts. The technology presents opportunities as well as, often ethical and legal, challenges, and has the potential for both positive and negative impacts for organisations, society, and individuals. Offering multi-disciplinary insight into some of these, this article brings together 43 contributions from experts in fields such as computer science, marketing, information systems, education, policy, hospitality and tourism, management, publishing, and nursing. The contributors acknowledge ChatGPT's capabilities to enhance productivity and suggest that it is likely to offer 1 Equal contributions Y.K. Dwivedi et al.International Journal of Information Management 71 (2023) 102642 3 significant gains in the banking, hospitality and tourism, and information technology industries, and enhance business activities, such as management and marketing. Nevertheless, they also consider its limitations, disruptions to practices, threats to privacy and security, and consequences of biases, misuse, and misinformation. However, opinion is split on whether ChatGPT's use should be restricted or legislated. Drawing on these contributions, the article identifies questions requiring further research across three thematic areas: knowledge, transparency, and ethics; digital transformation of organisations and societies; and teaching, learning, and scholarly research. The avenues for further research include: identifying skills, resources, and capabilities needed to handle generative AI; examining biases of generative AI attributable to training datasets and processes; exploring business and societal contexts best suited for generative AI implementation; determining optimal combinations of human and generative AI for various tasks; identifying ways to assess accuracy of text produced by generative AI; and uncovering the ethical and legal issues in using generative AI across different contexts. Collapse

Topics: artificial intelligence generative AI Twitter privacy innovation management

Methods: chatGPT chatbot generative pre-trained transformer literature study language model

Hey Article, What Are You About? Question Answering for Information Systems Articles through Transformer Models for Long Sequences

2023 | HICSS | Citations: 0

Authors: Ebert, L.; Huettemann, S.; Mueller, R.

Abstract: Question Answering (QA) systems can significantly reduce manual effort of searc ... Expand

Abstract: Question Answering (QA) systems can significantly reduce manual effort of searching for relevant information. However, challenges arise from a lack of domain-specificity and the fact that QA systems usually retrieve answers from short text passages instead of long scientific articles. We aim to address these challenges by (1) exploring the use of transformer models for long sequence processing, (2) performing domain adaptation for the Information Systems (IS) discipline and (3) developing novel techniques by performing domain adaptation in multiple training phases. Our models were pre-trained on a corpus of 2 million sentences retrieved from 3,463 articles from the Senior Scholars' Basket and fine-tuned on SQuAD and a manually created set of 500 QA pairs from the IS field.In six experiments, we tested two transfer learning techniques for fine-tuning (TANDA and FANDO). The results show that fine-tuning with task-specific domain knowledge considerably increases the models' Collapse

Topics: Wikipedia online learning system technology adoption IS discipline crowdsourcing

Methods: experiment machine learning deep learning computational algorithm transformer model

Theories: adaptive structuration theory structuration theory

Blockchain Transformation for Banking and Financial Services: Examining the Opportunities and Risks

2023 | Americas Conference on Information Systems | Citations: 0

Authors: El-Haddadeh, Ramzi; Weerakkody, Vishanth; Hindi, Nitham; Sivarajah, Uthayasankar

Abstract: The advancement of blockchain technologies has created new opportunities for bu ... Expand

Abstract: The advancement of blockchain technologies has created new opportunities for businesses to deliver new innovative offerings while improving their existing efficiencies to stay relevant and maintain the required competitive advantage. As such, blockchain technology has emerged as an opportunity to radically revolutionize and transform the banking and financial industry. This paper aims to bridge the existing gap in the literature by identifying empirically relevant opportunities and anticipated risks of blockchain transformation in the sector. This study collated and empirically analyzed the conceptualized model using a qualitative case study approach, culminating with specific recommendations. The findings highlight how the realization of blockchain technology value, competitive advantage, innovation and impact is required to develop successful implementation strategies. Similarly, acknowledging associated risks from scalability, Interoperability, security, and regulatory perspectives is equally vital to developing the necessary measures to guide its successful implementation. Collapse

Topics: blockchain systems implementation innovation management IS value competitive advantage

Methods: qualitative interview case study personal interview synthesis

z-Commerce: Designing a Data-Minimizing One-Click Checkout Solution

2023 | DESRIST | Citations: 0

Authors: Ermolaev, Egor; Abellán Álvarez, Iván; Sedlmeir, Johannes; Fridgen, Gilbert

Abstract: E-commerce has grown rapidly over the past years, with prevailing e-commerce pla ... Expand

Abstract: E-commerce has grown rapidly over the past years, with prevailing e-commerce platforms aggregating large amounts of customer data. This practice has several undesirable side effects, such as facilitating profiling that may lead to price discrimination and data feedback loops that can hamper competition. Moreover, data hoarding carries security risks through data breaches and undermines customers’ privacy expectations. On the other hand, convenience aspects and compliance regulation demand the processing and storage of user-related data. To address this tension field, we aim to conceptualize and iteratively refine a data-minimizinig e-commerce platform. Following a design science research approach, we identify design objectives and propose and implement a solution in which stakeholders receive only customer data that is indispensable for their part of the process. Our solution leverages digital identity wallets and general-purpose zero-knowledge proofs (zk-SNARKs). We aim to perform a criteria-based evaluation to assess our artifact’s feasibility and fitness from an interdisciplinary perspective. With our results, we hope to illustrate that combining state-of-the-art cryptographic techniques and an emerging digital identity paradigm allows reaching the user experience of incumbent e-commerce platforms while mitigating the undesirable socio-economic side effects of avoidable data disclosure. Collapse

Topics: electronic commerce privacy user experience open source price discrimination

Methods: qualitative interview design artifact design science literature study business process modeling

Exploring the Synergies in Human-AI Hybrids: A Longitudinal Analysis in Sales Forecasting

2023 | Americas Conference on Information Systems | Citations: 0

Authors: Fahse, Tobias; Schmitt, Anuschka

Abstract: Despite the promised potential of artificial intelligence (AI), insights into r ... Expand

Abstract: Despite the promised potential of artificial intelligence (AI), insights into real-life human-AI hybrids and their dynamics remain obscure. Based on digital trace data of over 1.4 million forecasting decisions over a 69-month period, we study the implications of an AI sales forecasting system's introduction in a bakery enterprise on decision-makers' overriding of the AI system and resulting hybrid performance. Decisionmakers quickly started to rely on AI forecasts, leading to lower forecast errors. Overall, human intervention deteriorated forecasting performance as overriding resulted in greater forecast error. The results confirm the notion that AI systems outperform humans in forecasting tasks. However, the results also indicate previously neglected, domain-specific implications: As the AI system aimed to reduce forecast error and thus overproduction, forecasting numbers decreased over time, and thereby also sales. We conclude that minimal forecast errors do not inevitably yield optimal business outcomes when detrimental human factors in decision-making are ignored. Collapse

Topics: artificial intelligence task productivity decision making IT productivity usability

Methods: digital trace data longitudinal research theory development ordinary least square qualitative interview

Theories: technology acceptance model socio technical theory

“Maybe You Should Talk to Someone”: The Role of Online Communities on Mental Healthcare

2023 | International Conference on Information Systems | Citations: 0

Authors: Ghosh, Soham; Chakraborty, Soumyakanti; Sharma, Megha; Susarla, Anjana

Abstract: Online Health Communities like YouTube offer mental health patients an alternat ... Expand

Abstract: Online Health Communities like YouTube offer mental health patients an alternative channel to learn about mental illnesses, the treatment path to follow, and to share their experiences. For many patients who are reluctant to seek professional help, a video on mental health uploaded by a content creator may serve as a substitute for a counsellor. Our work aims to develop an understanding of the relationship between language formality and social support and provide normative guidelines for content creators on social media platforms. Using two transformer-based deep learning classification models, we determine the degree of language formality or informality present in the content, and three dimensions of social support in the comments. We then utilize propensity score estimation to establish the causal effect of (in)formality on the dimensions of social support for 994 videos and 3,10,157 comments. Our findings indicate that informal speech increases emotional support, leading to better health outcomes. Collapse

Topics: YouTube social media electronic mail blog

Methods: deep learning RoBERTa BERT propensity score matching propensity score method

Theories: speech act theory

Mastering Agile Practice Adoption through a Model-Driven Approach for the Combination of Development Methods

2023 | Business & Information Systems Engineering | Citations: 0

Authors: Giachetti, Giovanni; de la Vara, José Luis; Marín, Beatriz

Abstract: Many software companies are adapting their traditional development processes to ... Expand

Abstract: Many software companies are adapting their traditional development processes to incorporate agile practices. In this context, it is necessary to count on expert knowledge to evaluate different agile practices and configure them according to project needs. However, this expert knowledge is scarce, difficult to validate, and time-consuming, since it is applied manually. As a solution, the paper presents a model-driven approach, called SIAM, which automatically generates guidelines for the adoption of agile practices through the combination of different development methods. SIAM is supported by a meta-model architecture to implement a knowledge repository that characterizes method configuration decisions, which can be reused in different development projects. SIAM has been implemented in a tool suite that facilitates the specification of models and the identification of issues during the definition of the development processes. The approach has been successfully applied to reconfigure an industrial development process with agile methods, showing that the effort required for tailoring agile practices according to organizational standards is considerably reduced. Collapse

Topics: agile software development Scrum knowledge repository system development method waterfall model

Methods: design methodology business process modeling case study metamodel design science