TABLE 1.

Different components of CRISP-DM model and the associated tasks.

CRISP Components	Tasks	Literature and Description
Business understanidng	–Define business objectives – Risk Assessment analysis –Cost and benefit analysis –Technical requirement analysis – Define data analysis objectives and project planning	In Sharma and Osei-Bryson (2009) a framework for implementing various business understanding tasks is presented and highlights dependencies between them. In Sharma and Osei-Bryson (2008); Rao et al. (2012) an organizational-ontology for business understanding is presented. In Nino et al. (2015) various aspects of business understanding and challenges related to big data are discussed
Data understanding and preparation	– Data extraction – Data description – Data quality estimation – Data selection for modeling Data cleaning and feature extraction – Data exploration	In Duch et al. (2004) rule-based data extraction and understanding is discussed. Uddin et al. (2014) discuss various characteristics of big data for its efficient applications. Karkouch et al. (2016); Qin et al. (2016) discuss data properties, life-cycle of data from internet of things (IoT) for maintaining data quality from IoT. Cichy and Rass (2019) reviews various comparisons that provide data quality frameworks from different areas, including industrial production. Hazen et al. (2014); Ardagna et al. (2018) discuss methods for data quality management, monitoring, and assessments. Steed et al. (2017); Zhou et al. (2019); Andrienko et al. (2020)) discuss visualization methods and challenges of manufacturing and big data. Stanula et al. (2018)) discuss guidelines for data selection for understanding business and data in manufacturing
Modeling and evaluation	– Model assumption and selection techniques for modeling, parameter selection – Feature engineering – Model testing, result visualization and analysis – Model evaluation and description – Other data and modeling issues affecting model performance	In Diez-Olivan et al. (2019), Vogl et al. (2019), Bertsimas and Kallus (2020) reviews of various models, models building and evaluation for descriptive, diagnostic, predictive, and prescriptive analysis in industrial production and manufacturing are presented
Deployment	– Model utility assessment – Model monitoring, maintenance and updates – Users response evaluation – Model evaluation for data understanding and business understanding	Issues of model deployment related to human-lefted data science and model safety are discussed in HUMAN-CENTERED Data Science and Model Safety