DataOps as a Prerequisite for the Next Level of Self-Service Analytics – Balancing User Agency and Central Control

2023 | International Conference on Business Informatics | Citations: 0

Authors: Baars, Henning

Abstract: The area of Business Intelligence and Analytics (BIA) has repeatedly oscillated ... Expand

Abstract: The area of Business Intelligence and Analytics (BIA) has repeatedly oscillated between more central, efficiency-oriented, professionalized approaches and decentral, agility-oriented, user-driven ones. This paper investigates whether and how to alleviate that tradeoff by combining an agility-oriented self-service analytics approach with the professionalization-driven DataOps concept: DataOps aims at transferring ideas from DevOps to the realm of analytics, namely a mutual integration of Development and Operations and a high degree of professionalization and automation. A case study with a series of interviews and a workshop led to insights into the viability of such a combination. These inspire a theoretical concept for capturing the economics behind the approaches that is considering the (opportunity) costs of the components "user agency" and "central control". The concept has been evaluated with representatives from the case study. Based on the results, I argue that the discussed combination can push BIA solutions towards fine-tuned federated environments. Collapse

Topics: devops artificial intelligence mlops data management data integration

Methods: qualitative interview case study literature study machine learning survey

Toward the Sustainable Development of Machine Learning Applications in Industry 4.0

2023 | European Conference On Information Systems | Citations: 0

Authors: Ellenrieder, Sara; Jourdan, Nicolas; Biegel, Tobias; Cassoli, Beatriz; Metternich, Joachim; Buxmann, Peter

Abstract: As the level of digitization in industrial environments increases, companies ar ... Expand

Abstract: As the level of digitization in industrial environments increases, companies are striving to improve efficiency and resilience to unplanned disruptions through the development of Machine Learning (ML)based applications. Still, sustainable deployment and operation beyond proofs-of-concept is a challenging and resource-intensive task in dynamic enviroments such as industry 4.0, often impeding practical adoption in the long-term and thus sustainable ML product development. In this work, we systematically identify these challenges based on the CRISP-ML process model phases by applying a design science research approach. To this end, we conducted 15 interviews with data science practitioners in industry 4.0. Following a qualitative content analysis, design requirements and design principles for the development and sustainable long-term deployment of ML systems are derived to address identified challenges such as robustness to, and management of data drift caused by timedependencies and machine/product differences, missing metadata, interfaces to other IT systems, expectation management, and MLOps guidelines. Collapse

Topics: production management industry 4.0 systems design systems development new product development

Methods: qualitative interview business process modeling design requirement design principle Cross-Industry Standard Process for Data Mining

INTEGRATING MACHINE LEARNING WITH SOFTWARE DEVELOPMENT LIFECYCLES: INSIGHTS FROM EXPERTS

2022 | European Conference On Information Systems | Citations: 0

Authors: Laato, Samuli; Mäntymäki, Matti; Minkkinen, Matti; Birkstedt, Teemu; Islam, A. K. M. Najmul; Dennehy, Denis

Abstract: This paper examines the challenges related to integrating machine learning (ML) ... Expand

Abstract: This paper examines the challenges related to integrating machine learning (ML) development with software development lifecycle (SDLC) models. Data-intensive development and use of ML are gaining popularity in information systems development (ISD). To date, there is little empirical research that explores the challenges that ISD practitioners encounter when integrating ML development with SDLC frameworks. In this work we conducted a series of expert interviews where we asked the informants to reflect upon how four different archetypal SDLC models support ML development. Three high level trends in ML systems development emerged from the analysis, namely, (1) redefining the prescribed roles and responsibilities within development work; (2) the SDLC as a frame for creating a shared understanding and commitment by management, customers, and software development teams: and (3) method tailoring. This study advances the body of knowledge on the integration of conceptual SDLC models and ML engineering. Collapse

Topics: systems development mlops devops artificial intelligence spiral development model

Methods: qualitative interview design methodology synthesis qualitative content analysis case study

Beyond MLOps: The Lifecycle of Machine Learning-based Solutions

2022 | Americas Conference on Information Systems | Citations: 0

Authors: Mucha, Tomasz; Ma, Sijia; Abhari, Kaveh

Abstract: Organizations increasingly use machine learning (ML) to transform their operati ... Expand

Abstract: Organizations increasingly use machine learning (ML) to transform their operations. The technical complexity and unique challenges of ML lead to the emergence of ML operations (MLOps) practices. However, the research on MLOps is in its infancy and is fragmented across disciplines. We extend and integrate these conversations by developing a framework that accounts for the technical, organizational, behavioral, and temporal aspects of the overarching ML-based solution lifecycle. We identify the key components of ML-based solution lifecycle and their configuration through an in-depth study of Finland's Artificial Intelligence Accelerator (FAIA) and follow-up semi-structured interviews with experts from multiple international organizations outside FAIA. This study contributes to the recent IS literature concerned with the sociotechnical aspects of ML. We bring new insights into the discussion on organizational learning, conjoined agency, and automation and augmentation. These insights extend and complement MLOps practices, thereby helping organizations better realize the potential of ML technology. KeywordsArtificial intelligence, machine learning, design, machine learning operations (MLOps), development operations (DevOps), lifecycle. Collapse

Topics: mlops organization learning artificial intelligence organizational context business process management

Methods: qualitative interview design artifact personal interview participant observation theory development

Theories: socio technical theory theory of abduction

From Ethical AI Principles to Governed AI

2021 | International Conference on Information Systems | Citations: 16

Authors: Seppälä, Akseli; Birkstedt, Teemu; Mäntymäki, Matti

Abstract: This study explores how organizations translate principles of ethical artificia ... Expand

Abstract: This study explores how organizations translate principles of ethical artificial intelligence (AI) into practice. To date, the research on AI ethics has been mostly conceptual, with a significant emphasis on defining the principles of ethical AI. Thus, there is momentum for a shift from principle-based ethics toward an increased focus on the implementation of ethical principles in practice. In this study, we analyzed data collected through a set of expert interviews in organizations deploying AI systems. We identified that ethical AI principles are implemented through four sets of practices: i) governance, ii) AI design and development, iii) competence and knowledge development, and iv) stakeholder communication. As our contribution to IS research, we empirically elucidate how the principles of ethical AI are translated into organizational practices. For IS practice, we provide organizations deploying AI with novel insights on how to govern their AI systems. Collapse

Topics: artificial intelligence responsible artificial intelligence mlops supply chain management website

Methods: qualitative interview computational algorithm qualitative coding

Interpretable Models & Metabehaviors: A Proposed Study of Microlending

2021 | Americas Conference on Information Systems | Citations: 0

Authors: Slaughter, Kelly; Preston, David

Abstract: As AI becomes embedded in everyday professional and personal tasks, the need fo ... Expand

Abstract: As AI becomes embedded in everyday professional and personal tasks, the need for accessible and understandable explanations for AI decisions is compelling. Accordingly, firms are developing solutions (either voluntarily or through government mandate) to provide explanations of AI decision-making to those who are subject the consequences of the decisions. Given the recency of this progress, limited research has been conducted as to how the introduction of these explanations may alter the relationship between humans, agents, model findings, and outcomes. First, humans have a long history of "outsmarting" systems that affect their welfare. Second, humans are not passive recipients of the explanations for model decisionmaking but will incorporate new information into future actions. Our research interest is examining how humans come to evaluate AI decision-making without explicit explanations and how contingencies in explanations alter the human's behavior and model operations. Collapse

Topics: artificial intelligence decision making open source shapley additive explanation mlops

Methods: experimental group laboratory experiment longitudinal research PyTorch machine learning