OBIDAM KT 1-3
: Which of the following statements are true according to Graham (2015) in "Chapter 2: What is Master Data":
- "Reference data" have the following characteristics: slowly changing, simple in structure and have often few rows and columns. - Transaction structure data" are needed to give transactions some meaning. - "Enterprise structure data" describe hierarchies in an organization, such as business functions and product grouping - and each node in the hierarchy is a "transaction structure data" or a "reference data".
Which of the following statements are true regarding data profiling, according to Plotkin (2014) in "Chapter 7: Important Roles of Data Stewards":
- . A data profiling tool can be used to measure the conformance to data quality rules - also called "goodness to fit". - A data profiling tool can be used to discover possible data quality rules.
: Which are major reasons for low data quality, according to Plotkin (2014) in "Chapter 7: Important Roles of Data Stewards
- . Data quality rules are not enforced at the point of data origins and during data loads - Data users are making their own corrections when identifying data quality issues, instead of informing a responsible for data quality (for example, a data steward that will enhance the data quality also for downstream users). - Data producers are not accurate and complete when producing data, for example, they may leave blanks and do not change default values when needed. - Data quality rules are not defined.
Which of the following statements are true regarding subject-based classification according to the article by Garshol (2004) "Metadata? Thesauri? Taxonomies? Topic Maps Making sense of it all":
- A controlled vocabulary is a closed list of named subjects, which can be used for classification. - Classification is about grouping objects in classes, and a subject-based classification groups objects based on subjects (the objects are about). - Metadata describe objects and one way of doing that is by classifying objects by using the subjects that the objects are about - Metadata can be used for information retrieval - There is a difference in describing the objects being classified, and describing the subjects used to classify the objects. Metadata describe objects and one way of doing that is by classifying objects by using the subjects that the objects are about - An ontology is a model for describing the world that consists of a set of types, properties of the types and relationships types. - A taxonomy describes the subjects being used for classification, but it is not itself metadata, it can, however, be used in metadata. - The idea of classifying objects using a facet classification is to select one term from each facet to describe the object. Facets can be thought of as different axes along which an object can be classified.
Which of the following statements are true according to the article: McAfee and Brynjolfsson (2012): "Big Data: The Management Revolution":
- A powerful approach for changing a decision-making culture towards a more data-driven one is to see executives allowing themselves to be overruled by the data, instead of trusting their intuitions if data contradict the intuitions. - The difference between traditional analytics and big data are: Volume, Velocity and Variety. - The case from Sears in the article shows that one important reason for the time dropped for generating a comprehensive set of promotions from eight weeks to one was to avoid transferring data between data sources and combining them, which is time consuming - An investigation with 330 North American companies showed that companies in the top third in their industries in their use of data-driven decision making, showed that they were, on average, 5 % more productive and 6 % more profitable.
Which of the following statements are true according to Plotkin (2014) in "Chapter 2: Understanding the Types of Data Stewardship":
- A technical data steward is an IT personal that understand how data are created, stored and moved in technical systems in the organization. - A domain data steward is an individual that steward data elements that are shared in several business areas.
Which of the following steps are part of the 6-step engineering process presented in Chapter 10 (book excerpt) in Schmarzo (2013) Big Data. Understanding How Data Power Big Business:
- Brainstorm big data business impact. - Identifying your organization's key business initiatives. - Understand how the organization makes money. - Design and implement big data solution. - Break down the business initiative into use cases.
For each use case that target a business initiative, the following info need to be captured, according to Chapter 10 (book excerpt) in Schmarzo (2013) Big Data. Understanding How Data Power Big Business
- Capture data needed, that is, requirements on data (i.e. data sources, key metrics, level of granularity, frequency of access, hierarchies etc). - Capture business questions that the business stakeholders need to ask - if they have access to more data sources and more detailed data - Capture use experience requirements (that are closely connected with the user's decision making process). - Capture business decision that the business stakeholders need to make. - Capture analytic algorithms and modelling needed, that is, requirements on analytic algorithms and modelling to be applied. - Capture business stakeholders to be targeted.
: Which of the following statements are true regarding subject-based classification according to the article by Garshol (2004) "Metadata? Thesauri? Taxonomies? Topic Maps Making sense of it all:
- Categories are terms in a subject-based classification. Therefore, a set of categories can be seen as a controlled vocabulary, and, if the categories in the set are hierarchical arranged, it can be seen as a taxonomy. - A taxonomy is about arranging the terms in a controlled vocabulary into a hierarchy
Which of the following data are examples of master data, according to Chapter 1 (book excerpt) in Loshin (2009) Master Data Management:
- Employees. - Locations. - Products. - Policies. - Contracts. - Customers. - Suppliers.
: Which of the following statements are true according to the article: Davenport et al (2012): "How Big Data are different"
- Examples of approaches, techniques and products for managing big data are Hadoop, the Cloud, and Virtual data marts.
Which of the following statements are true according to Plotkin (2014) in "Chapter 1: Data Stewardship and Data Governance: How They Fit Together
- Governed data include, for example, that the business data elements are mapped to physical data locations. - Data governors often represent business functions for data-related matters. Data - Data stewardship is the operational aspects of data governance. - Governed data include, for example, that usage rules for data elements are specified. - A data warehouse implementation could be a driver for getting data into a governed state.
Which of the following statements are true according to Graham (2015) in "Chapter 3: The Business Case for MDM":
- In a Merger and Acquisition (M&A) implementation, the possibility to hook up to an existing MDM hub is an important enabler for saving cost during the M&A implementation - DM is important for being successful with CRM - MDM increase the ability to comply with regulatory compliance requirements
: What is true about master data management (MDM) system approaches, according to the paper White (2007) Using master data in business intelligence
- In the centralized approach, the MDM systems work as both the systems of entry (SOE) and system of record (SOR) for the master data. - The differences between a consolidate approach and a propagate approach is that the former consolidate operational master data into a single master data store that becomes the system of record (SOR) for master data for all systems of entry (SOE), while a propagate approach means that data in all operational system, that is, the systems of entry (SOE), are kept consistence.
The following statement are true regarding the step 5: Prove out the use case, according to Chapter 10 (book excerpt) in Schmarzo (2013) Big Data. Understanding How Data Power Big Business:
- In this step, data transformation processes need to be defined, including, cleansing, aligning, preparations, enrichment of data. - The step aims to validate that the data and analytics can deliver meaningful and actionable insights and recommendations for the targeted business solution. - In this step, mock-ups and/or wireframes could be developed that will help business stakeholders to understand how the resulting analytic model will be integrated into their dailybusiness processes.
Which of the following four ways can big data and predictive analytics power the organizations key business initiatives according to Chapter 10 (book excerpt) in Schmarzo (2013) Big Data. Understanding How Data Power Big Business
- Integrate unstructured data sources (such as consumer notes, call center notes, and social media posts) to the existing sources in order to include new variables and dimensions to the analysis, and, thereby, enable accurate decisions based on complete data. - Reduce time delay between data events and the analysis of the events since this will enable more frequent and timely decision. This could create real time location based insights and on-demand customer segments. - Mine transactional (dark) data, such as POS, RFID, Credit Card, at the lowest level of granularity in order to uncover new opportunities for the business initiatives. This will enable more detailed decisions. - . Integrate predictive analysis into your business processes to uncover causality buried inthe data, and, thereby, enable more actionable and predictive decisions.
: What differs MDM initiatives of today from previous attempts to consolidate data, such as data warehousing (DW), business intelligence (BI), customer relationship management (CRM), according to Chapter 1 (book excerpt) in Loshin (2009) Master Data Management:
- MDM aims to provide methods for managed access to a unified view of enterprise data objects, and not for creating another silo of enterprise data copies. - The previous attempts have often been technology driven.
Which of the statements are true for a master data management system and the components of such system, according to the paper White (2007) Using master data in business intelligence
- Master data integration services - consolidate and propagate master data. - Master metadata store - contains metadata in form of a data model (with business entities, their attributes and relationships), master data rules and definitions. In this store, information about SOR is documented. - MDM applications - manage and publish master data and metadata. - In a fully compliant MDM system, the MDM system is also the SOE for all master data.
Which of the following statements regarding master data are true, according to Chapter 1 (book excerpt) in Loshin (2009) Master Data Management:
- Master data tend to exist in more than one business area.
Which of the following statements are true according to the article: Ross et al (2012): "You May Not Need Big Data after All ":
- One problem with the promises of big data is that analytics-generated insights can be easy to replicate by other companies. - The Seven Eleven Japan case showed the importance of coaching employees to make decisions based on data. - A way of handling the complexity of business rules is to embed them in software -thereby it frees employees from routine decisions. - The Seven Eleven Japan case presented in the article is an example of an approach to generate big value from little amount of data by, for example, in this case empowering the salesclerks with up-to date sales data and possibilities to change the orders to suppliers on a daily basis - One problem with the promises of big data is that turning insights from data analytics into competitive advantage requires changes in business that organizations may be incapable to do - According to an investigation based on seven case studies and interview with executives at 51 organizations, organizations that have a culture of evidence based decision making tend to be more profitable then organizations lacking that culture. - The biggest reason for failing paying of investments in big data is that organizations do not know how to do a good job with information
How to create a data-driven organization with a culture of evidence-based decision-making, according to the article: Ross et al (2012):? Which of the alternatives represent useful suggestions?
- Provide (near) real-time feedback to decisions made. - Agree on a Single Source Of Truth. - Manage business rules explicitly. - Improve performance by active coaching
: Which of the following statements are true according to Plotkin (2014) in "Chapter 7: Important Roles of Data Stewards":
- Reference data can be represented as an enumerated attribute which list valid values. - Reference data are data that include product types, status codes and list of valid values, and that are referred to by systems, applications, processes and other data records. - Reference data are often created by organizations to standardize their own information. - One important task for data stewards is to make incentives for data producers to be accurate and complete, for example, do not left blanks, even if the accurateness and completeness do not directly benefit the data producers themselves. - Data quality needs to be measured against a set of requirements (or data quality rules) and these requirements are based on how the business wants to use the data. - Data quality rules should consist of two parts: 1) a statement explaining what quality means in business terms, and 2) a specification explaining what is good quality at the physical database level.
Which of the following statements are true regarding data quality dimensions according to the book excerpt by Plotkin (2014) " Chapter 7: Important Roles of Data Stewards":
- The data quality dimension "accuracy" means the degree to which data correspond to known correct values in the real world, provided by a recognized source of truth. - The data quality dimension "integrity" means the degree to which data contain consistent content across multiple databases.
Which of the following statements are true regarding a master data management (MDM) program, according to Chapter 1 (book excerpt) in Loshin (2009) Master Data Management:
- The intention of a MDM program is to create a single repository of high quality master data that feed applications across the organization with a consistent view of enterprise data. - A directive for succeeding with a MDM program is to manage the metadata properly.
Which of the following statements are true according to the article: McAfee and Brynjolfsson (2012): "Big Data: The Management Revolution":
- To be an effective organization, information and decision rights have to be put in the same location
Which of the following statements are true according to the article: Davenport et al (2012) "How Big Data are different":
Big data makes it possible to make personalized offers in milliseconds based on analysis of streaming data, and this requires changes in processes and decision making - Identifying and developing new products and services are an important part of a data scientists work - Big data is, for example, about real-time monitoring
Which of the following statements are true according to Plotkin (2014) in "Chapter 6: Practical Data Stewardship":
Financial reporting data are data elements that need to governed, as well as compliance and regulatory data elements. - A business glossary can warn for potential duplicates. - The starting point for a data stewardship is to identify the key business elements, and then assigning stewardship for them. - An important task for data stewardship is to define creation and usage business rules. - A metadata repository usually focused on physical and technical metadata but business metadata can also be part of the repository. - A business glossary include business metadata, such as data definitions, data quality requirements, creation and usage business rules and info of responsible data stewards
Which of the following statements are true according to the article "What is Your Data Strategy?" by DalleMule & Davenport (2017):
In few large organizations you can find a data strategy where data is both tightly controlled and used in a flexible way. - An important approach for developing a sound data strategy, is that organizations derive multiple versions of the truth from the same single source of truth.
: Which of the following statements are true, according to the paper White (2007) Using master data in business intelligence
In most companies' master data is maintained by multiple systems of entry (SOE). - A system of entry (SOE) is the system where master data is created and maintained. - A system of record (SOR) is the gold copy where master data is guaranteed to be accurate and up-to-date. - A main objective of a MDM system is to publish master data that have the quality to be used by applications and users. - MDM requires not only applications and technologies, but also an organization implementing policies and processes for controlling how master data are created and used - The new things with MDM compared with earlier attempt handling master data in transactional systems and data warehousing system is that 1) master data need to be managed and integrated outside existing systems, 2) vendors provide packaged solutions, i.e. applications and technologies, supporting MDM
