What Makes Digital Technology? A Categorization Based on Purpose

2023 | Communications of the Association for Information Systems | Citations: 0

Authors: Baier, Marie-Sophie; Berger, Stephan; Kreuzer, Thomas; Oberländer, Anna; Röglinger, Maximilian

Abstract: Digital technology (DT) is creating and shaping today's world. Building on its ... Expand

Abstract: Digital technology (DT) is creating and shaping today's world. Building on its identity and history of technology research, the Information Systems discipline is at the forefront of understanding the nature of DT and related phenomena. Understanding the nature of DT requires understanding its purposes. Because of the growing number of DTs, these purposes are diversifying, and further examination is needed. To that end, we followed an organizational systematics paradigm and present a taxonomic theory for DT that enables its classification through its diverse purposes. The taxonomic theory comprises a multi-layer taxonomy of DT and purpose-related archetypes, which we inferred from a sample of 92 real-world DTs. In our empirical evaluation, we assessed the reliability, validity, and usefulness of the taxonomy and archetypes. The taxonomic theory exceeds existing technology classifications by being the first that (1) has been rigorously developed, (2) considers the nature of DT, (3) is sufficiently concrete to reflect the diverse purposes of DT, and (4) is sufficiently abstract to be persistent. Our findings add to the descriptive knowledge on DT, advance our understanding of the diverse purposes of DT, and lay the ground for further theorizing. Our work also supports practitioners in managing and designing DTs. Collapse

Topics: knowledge base perceived usefulness competitive advantage decision making technology adoption

Methods: cluster analysis theory development longitudinal data analysis focus group computational algorithm

Theories: information systems theory sociomaterialism theory

How Virtuous are Virtual Influencers? – A Qualitative Analysis of Virtual Actors’ Virtues on Instagram

2023 | HICSS | Citations: 0

Authors: Hofeditz, Lennart; Erle, Lukas; Timm, Lara; Mirbabaie, Milad

Abstract: Recently, virtual influencers (VIs) have become a more frequent alternative to h ... Expand

Abstract: Recently, virtual influencers (VIs) have become a more frequent alternative to human influencers (HIs). VIs can be described as non-human agents who behave in a human-like pattern. Big enterprises such as Prada, Porsche, Samsung, or Ikea have already collaborated with VIs in the past. Even though it should be clear to users that VIs cannot practice values and virtues in the real world, VIs seem to express certain virtues. This research paper focuses on identifying virtues conveyed by VIs and the effect of expressing virtues on follower engagement by conducting a qualitative content analysis of social media posts. Furthermore, we checked on VIs being abused by companies to convey a more favorable image. Our findings suggest that conveying certain virtues seems to have a positive effect on the engagement. In addition, some VIs were used by companies for virtue signaling without being noticed by their followers. Collapse

Topics: social media marketing management Instagram advertising management database system

Methods: qualitative content analysis survival analysis computational algorithm

Building a deep learning-based QA system from a CQA dataset

2023 | Decision Support Systems | Citations: 0

Authors: Jin, Sol; Lian, Xu; Jung, Hanearl; Park, Jinsoo; Suh, Jihae

Abstract: A man-made machine-reading comprehension (MRC) dataset is necessary to train the ... Expand

Abstract: A man-made machine-reading comprehension (MRC) dataset is necessary to train the answer extraction part of existing Question Answering (QA) systems. However, a high-quality and well-structured dataset with questionparagraph-answer pairs is not usually found in the real world. Furthermore, updating or building an MRC dataset is a challenging and costly affair. To address these shortcomings, we propose a QA system that uses a large-scale English Community Question Answering (CQA) dataset (i.e., Stack Exchange) composed of 3,081,834 questionanswer pairs. The QA system adopts a classifier-retriever-summarizer structure design. The question classifier and the answer retriever part are based on a Bidirectional Encoder Representations from Transformers (BERT) Natural Language Processing (NLP) model by Google, and the summarizer part introduces a deep learning-based Text-to-Text Transfer Transformer (T5) model to summarize the long answers. We instantiated the proposed QA system with 140 topics from the CQA dataset (including topics such as biology, law, politics, etc.) and conducted human and automatic evaluations. Our system presented encouraging results, considering that it provides highquality answers to the questions in the test set and satisfied the requirements to develop a QA system without MRC datasets. Our results show the potential of building automatic and high-performance QA systems without being limited by man-made datasets, a significant step forward in the research of open-domain or specificdomain QA systems. Collapse

Topics: privacy decision making online community IT productivity Google+

Methods: BERT chatGPT computational algorithm experiment natural language processing

Illuminating Smart City Solutions – A Taxonomy and Clusters

2023 | International Conference on Information Systems | Citations: 0

Authors: Jonas, Claudius; Schmitt, Kevin; Oberländer, Anna Maria; Ebel, Philipp

Abstract: With urban problems intensifying, Smart City solutions are recognized by resear ... Expand

Abstract: With urban problems intensifying, Smart City solutions are recognized by researchers and practitioners as one of the most promising solutions to make urban areas economically, environmentally, and socially sustainable. While many elements of Smart City solutions have been explored, existing works either treat Smart City solutions as technical black boxes or focus exclusively on Smart City solutions' technical or nontechnical characteristics. Therefore, to conceptualize the unique characteristics of Smart City solutions currently available, we developed a multi-layer taxonomy based on Smart City solution literature and a sample of 106 Smart City solutions. Moreover, we identified three clusters, each covering a typical combination of characteristics of Smart City solutions. We evaluated our findings by applying the Q-sort method. The results contribute to the descriptive knowledge of Smart City solutions as a first step for a theory for analyzing and enable researchers and practitioners to understand Smart City solutions more holistically. Collapse

Topics: smart city internet of things mobile system mobile application analytical information system

Methods: survey computational algorithm cluster analysis design process design methodology

Detecting Feature Requests of Third-Party Developers through Machine Learning: A Case Study of the SAP Community

2023 | HICSS | Citations: 0

Authors: Kauschinger, M.; Vieth, N.; Schreieck, M.; Krcmar, H.

Abstract: The elicitation of requirements is central for the development of successful so ... Expand

Abstract: The elicitation of requirements is central for the development of successful software products. While traditional requirement elicitation techniques such as user interviews are highly labor-intensive, data-driven elicitation techniques promise enhanced scalability through the exploitation of new data sources like app store reviews or social media posts. For enterprise software vendors, requirements elicitation remains challenging because app store reviews are scarce and vendors have no direct access to users. Against this background, we investigate whether enterprise software vendors can elicit requirements from their sponsored developer communities through data-driven techniques. Following the design science methodology, we collected data from the SAP Community and developed a supervised machine learning classifier, which automatically detects feature requests of third-party developers. Based on a manually labeled data set of 1,500 questions, our classifier reached a high accuracy of 0.819. Our findings reveal that supervised machine learning models are an effective means for the identification of feature requests. Collapse

Topics: enterprise information system requirements elicitation requirements analysis Fiverr cloud computing

Methods: machine learning supervised learning design science naïve bayes classification method

Tools of Trade of the Next Blue-Collar Job? Antecedents, Design Features, and Outcomes of Interactive Labeling Systems

2023 | European Conference On Information Systems | Citations: 0

Authors: Knaeble, Merlin; Nadj, Mario; Germann, Luisa; Maedche, Alexander

Abstract: Supervised machine learning is becoming increasingly popular -and so is the nee ... Expand

Abstract: Supervised machine learning is becoming increasingly popular -and so is the need for annotated training data. Such data often needs to be manually labeled by human workers, not unlikely to negatively impact the involved workforce. To alleviate this issue, a new information systems class has emerged -interactive labeling systems. However, this young, but rapidly growing field lacks guidance and structure regarding the design of such systems. Against this backdrop, this paper describes antecedents, design features, and outcomes of interactive labeling systems. We perform a systematic literature review, identifying 188 relevant articles. Our results are presented as a morphological box with 14 dimensions, which we evaluate using card sorting. By additionally offering this box as a web-based artifact, we provide actionable guidance for interactive labeling system development for scholars and practitioners. Lastly, we discuss imbalances in the article distribution of our morphological box and suggest future work directions.To alleviate these issues, labeling tasks are nowadays performed in cooperation with machine intelligence. Hereby, a new class of information systems has emerged -that of interactive labeling (IL) systems. IL is defined as "combining automatic steps with incremental user input" (Bernard et al., 2018). IL differentiates from plain labeling tasks by incorporating machine intelligence in an interactive fashion, as opposed to using manual work start to end. For instance, when having crowdworkers annotate documents, Ramos et al. (2020) provide guidance in form of an iterative, incremental process, leveraging the workers capabilities to teach humans a skill. With only minimal training, their novices matched the performance of professionals. Their guidance further leads to lower frustration among participants. Chang et al. (2017) have worked on designing IL systems leveraging the scalability of crowdworking platforms. When having images labeled, they used a series of synchronized stages involving multiple workers each. By allowing for ambiguity and labeling as structures instead of binary decisions, they could decrease instruction effort and allow for an ex-post setting of label boundaries. While it may be easy to decide that a photo of a house cat should be labeled 'cat', and that of a Caterpillar-branded (short CAT) excavator should not be, the decision for a wild leopard is not straight forward. Hereby the stages of their system allow for labeler-led creation of new subcategories, if disagreement occurs. This is especially helpful to prevent fatal inconsistencies in the dataset, if, for instance, some labelers find leopards to be cats, while others do not. The IL system asks labelers for explanations for such disagreement, extracts sub-category names, and uses these as new labels to allow for the task issuers to resolve the question after labeling has concluded. So far, a considerable number of articles have already been published despite the young nature of this new class of information systems. Amershi et al. (2014) provide an initial overview, differentiate IL from previous works, and introduce the mistreatment of users as oracles. Further, we have previously studied user archetypes addressed in IL research illustrating two competing perspectives of how users are treated (Knaeble et al., 2020): (a) oracle (e.g. users are asked if a label is correct) versus (b) teacher (e.g. users can provide deeper explanations in labeling tasks). Moreover, recently, we derived five general design principles along prominent literature findings and exemplary IL approaches illustrating that the design of such systems is diverse, and can significantly vary in scope and functionality . However, this diversity encompasses a complex of dimensions and characteristics across different IL systems that is not currently well understood. Y. Zhang et al. (2022) have recently argued for more support for those developing IL systems, as they commonly need to be tailor-made. Knowing which design features IL systems offer (and which alternatives exist) is therefore of utmost importance. Hereby, designers and developers of IL systems require an overview of the available design features of IL systems, to be able to make educated decisions. Furthermore, research needs to draw connections between such design features, foregoing antecedents, and following outcomes. Recent research in crowdworking has highlighted the importance of both (Alpar and Osterbrink, 2020). Antecedents are hereby task and context specific requirements and necessities, in the context of IL systems this could be what data is supposed to be labeled, whereas outcomes are goals which are seen as desirable by designers, such as efficiency. By understanding such antecedents and outcomes, IL designers and developers get examples on how to optimize their systems regarding given antecedents or desired outcomes. As IL is a young and popular field of research the need for conducting such a study is exacerbated further. As we will show, the amount of relevant research is increasing rapidly. While we have derived knowledge about how IL systems currently treat users (Knaeble et al., 2020) and we outlined general goals to aim for when creating IL systems , there is a gap of design knowledge on how to achieve such goals. Specifically there is no overview of existing design features, foregoing antecedents, and following outcomes. Hereby, research has created, implemented, and evaluated numerous design features of IL systems, but is yet to aggregate such results for a holistic picture. On this basis, we formulate the following research question (RQ): How to describe antecedents, design features, and outcomes of interactive labeling systems in a morphological box? Hereby, we rely on a systematic literature review (SLR), deductive and inductive coding, as well as a card sorting evaluation. Our theoretical contributions lie in the following. First, we present a literature-grounded and userevaluated systematic overview of a new class of information systems, along antecedents, design features, Collapse

Topics: database system recommender system crowdsourcing open source systems development

Methods: qualitative interview card sorting literature sample literature study qualitative coding

Between Hype, Hope, and Reality: A Lifecycle-Driven Perspective on Non-Fungible Token

2023 | European Conference On Information Systems | Citations: 0

Authors: Kölbel, Tobias; Joussen, Katrin; Weinhardt, Christof

Abstract: Advocates consider NFTs a potentially disruptive blockchain-enabled innovation. ... Expand

Abstract: Advocates consider NFTs a potentially disruptive blockchain-enabled innovation. In light of surging popularity and low theoretical insights, we study NFTs from a lifecycle-driven perspective. We develop a taxonomy that adheres to a habitual method and draws on a five-step process of analyzing literature and real-world projects. Our taxonomy contributes to descriptive knowledge by structuring NFTs with 20 dimensions and 77 characteristics along the perspectives of origination, distribution, transfer, trade, and redeem. We enable researchers and practitioners to grasp the NFT phenomenon in a structured manner and demonstrate the applicability of our taxonomy through expert interviews and case studies. Collapse

Topics: database system blockchain non-fungible token social network virtual reality

Methods: qualitative interview literature study theory development mixed method personal interview

Differential Impacts of Technology-Network Structures on Cost Efficiency: Knowledge Spillovers in Healthcare

2023 | Journal of Management Information Systems | Citations: 0

Authors: Nwafor, Onyi; Ma, Xiao; Johnson, Norman A.; Singh, Rahul; Aron, Ravi

Abstract: We examine how hospital cost efficiency can improve because of knowledge spillov ... Expand

Abstract: We examine how hospital cost efficiency can improve because of knowledge spillover effects, arising from the experiences of members of a healthcare system. We identify two types of functional technology networks—that we term repository-type and workflow-type networks—which are defined by the adoption patterns of different types of electronic health record (EHR) systems application functions by hospitals in a health system. Additionally, we examine how structural features of these two network types—namely, network centrality and interconnectedness—impact such knowledge spillover effects. By analyzing seven years of panel data for 1,420 U.S. hospitals across 216 health systems, and data obtained from surveys of healthcare experts, we found that greater centrality of nodes in repository-type networks enhances the influence of partners' experiences on cost efficiency but has the opposite effect in workflow-type networks, while greater interconnectedness of workflow-type networks enhances the influence of partners' experiences on cost efficiency. Interestingly, we found the opposite effect in repository-type networks, where greater interconnectedness diminishes the influence of partners' experiences on cost efficiency. We discuss reasons for this surprising and counterintuitive finding. Overall, our results demonstrate that at least in healthcare settings where information technologies enable knowledge-sharing networks to be formed, benefits can accrue through knowledge spillover effects; moreover, variations across functional categories of technologies adopted in a network context can result in different technology-network structures with differential impacts on hospital cost efficiency because of their influence on knowledge spillover at both the hospital and system levels. Not only can these results explain mixed findings in the literature that has been dominated by a focus on cost efficiency outcomes dependent solely on the EHR adoption choices of individual hospitals, but they also provide the basis for guidelines for a network-based perspective of technology adoption. We discuss other possible settings to which our empirical findings can be reasonably generalized, including supply chain technologies and the Internet of Things. Collapse

Topics: health information system knowledge sharing knowledge spillover technology adoption IT manager

Methods: longitudinal research robustness check descriptive statistic parametric test ANOVA

Theories: social network theory

Differential Effects of Multidimensional Review Evaluations on Product Sales for Mainstream vs. Niche Products

2023 | Management Information Systems Quarterly | Citations: 0

Authors: Zheng, Xin; Cao, Jisu; Hong, Yili; Yang, Sha; Ren, Xingyao

Abstract: Despite a large body of literature on online reviews, none have considered the ... Expand

Abstract: Despite a large body of literature on online reviews, none have considered the nuanced impacts of how multidimensional reviews affect product sales differently for mainstream vs. niche products. This study seeks to fill this knowledge gap by conducting complementary studies in two product categories (i.e., automobiles and laptops) with different methods (a field study and three lab experiments). Our paper reveals three key insights into the emerging literature and phenomenon on multidimensional review systems: (1) the interdimensional rating variance is more negatively related to product sales for mainstream products than for niche products in the same category, (2) the intradimensional rating valence on the dominant dimension of a product is more positively related to product sales for niche products than for mainstream products, and (3) the intradimensional rating variance on the dominant dimension of a product is more negatively related to product sales for niche products than for mainstream products. Our research provides important managerial implications for both product providers and review platforms. Collapse

Topics: laptop computer online review user-generated content big data marketing management

Methods: experiment field study analysis of covariance qualitative interview laboratory experiment

Competitive Poaching in Search Advertising: Two Randomized Field Experiments

2022 | Information Systems Research | Citations: 0

Authors: Bhattacharya, Siddharth; Gong, Jing; Wattal, Sunil

Abstract: Keyword searches with brand names occur very commonly on search engines, enablin ... Expand

Abstract: Keyword searches with brand names occur very commonly on search engines, enabling firms to infer not only when consumers are searching for them but also when consumers are searching for their competitors. This enables firms to generate traffic from search advertising by bidding not only on their own keywords but also on competitors’ keywords. The strategy of bidding on competitors’ keywords, known as competitive poaching, presents unique opportunities for academic research. In this research, we examine the factors that influence the effectiveness of competitive poaching, specifically the role of different ad copies and the type of competitor (poached brand) from which a brand is poaching. We also examine how the presence of sponsored ads from the poached brand affects competitive poaching. We collected data from two randomized field experiments, one with a business school in the Northeastern United States and the other one with a leading automobile dealership company, where these firms bid on keywords of competing brands and randomly display different types of ad copies in the sponsored search listings. We find that, when poaching on keywords of high-quality brands, ad copies that feature vertical differentiation are more effective than the control ad copies, which do not convey any differentiation or prescriptive messages. We also find that when poaching from low-quality brands, ad copies featuring horizontal differentiation perform better than the control ad copies. Finally, we show that the presence of the poached brand’s own ad has a positive association with the ad effectiveness of the poaching brand when that poached brand is high quality and a negative association when the poached brand is low quality. The academic and managerial contributions are also discussed. Collapse

Topics: search engine search engine marketing vertical differentiation horizontal differentiation Google+

Methods: field experiment chi squared test experiment randomized field experiment Poisson regression