| Microsoft SQL Server 2014 Business Intelligence Development: Beginners Guide |
Page 2 of 2
Author: Reza Rad
Chapter 5 Master Data Management Master Data Management (MDM) ensures only one copy of definitive data is kept. The chapter discusses the typical situation were various databases record similar information, but are often out-of-synch. Having a central database of master data removes this data inconsistency. Next, MDS is discussed, this provides features to enact MDM, including validation, security, business rules, and integration with DQS. The chapter continues with practical exercises/walkthroughs for:
This chapter provides an overview of MDS which ensures definitive data is stored once only.
Chapter 6 Data Quality and Data Cleansing Related data from different sources can introduce data quality issues. The chapter opens with a look at the need for high-quality data, to ensure algorithms match data better, leading to better decision making. Data problems discussed include: uniqueness, accuracy, and completeness. Next, the chapter looks at Data Quality Services (DQS), a tool for keeping data clean, based on a Knowledge Base, where data stewards create and maintain policies. The various components of DQS (cleansing, matching, monitoring, profiling) are discussed. The chapter continues with practical exercises/walkthroughs for:
This chapter provides an overview of the importance of clean data, some of the tools and methods for keeping data clean, and the integration with MDS and SSIS.
Chapter 7 Data Mining – Descriptive Models in SSAS Descriptive models analyse existing data, without making predictions. The chapter opens with an overview of data mining, an iterative cycle involving: problem definition, data preparation, data exploration, building and validating models, deploying and updating models. Next, 9 data mining algorithms are briefly explained, namely: Decision tree, Association rule, Clustering, Naive Bayes, Linear regression, Neural network, Logistic regression, Sequence clustering and Time series. Each algorithm has tuneable parameters to help get better results (e.g. set the number of clusters). The chapter continues with a detailed example on the use of the decision tree algorithm. This identifies nodes and builds trees on those nodes, the nodes are based on decisions on a set of attributes (e.g. age over 40). A useful step-by-step example is given, which builds nodes and trees to identify the attributes that identify buyers of bicycles. The chapter next provides another useful step-by-step example, this time using the association rules algorithm. This identifies relationships/rules between different variables in a dataset. The aim is to identify products commonly purchased together, which can be useful in cross-selling. This chapter provides a useful and practical introduction to data mining with descriptive models in SSAS. However, some terms are used before being defined (e.g. training and test case). Also, it might have been useful to link it to the wider realm of Machine Learning and Big Data. Chapter 8 Identifying Data Patterns – Predictive Models in SSAS Here the focus moves to identifying data patterns to make predictions, allowing better business decisions. The chapter opens with a look at finding the most suitable algorithm for a dataset, luckily Microsoft provides tools for this (e.g. Lift Chart and Profit Chart) – which compare the results from the different mining models. The chapter provides a step-by-step walkthrough of finding which is the better model (Naive Bayes or Clustering) when targeting a mail shot. The chapter continues with a look at using the Data Mining Extension (DMX) language to query and work with data mining algorithms. In the example given, a dataset of new customers is queried, to identify prospective bicycle buyers. The chapter ends with another step-by-step example, using the Time Series mining model, useful for predicting time-based facts (e.g. expected sales in 18 months time). This chapter provides a practical introduction to data mining with predictive models in SSAS. Chapter 9 Reporting Services SQL Server Reporting Services (SSRS) helps users visualize data as reports. The chapter opens with a high-level overview of the reporting services architecture, highlighting: SSRS databases, SSRS Web Applications, and SSRS Reports. The chapter continues with practical exercises/walkthroughs for:
This chapter provides a useful practical introduction to using SSRS, with helpful examples.
Chapter 10 Dashboard Design Microsoft has various tools for visualization, including SSRS, PerformancePoint, and Excel. This chapter focuses on using PerformancePoint to build dashboards and charts, and Power View to build charts. These tools have the advantage they are more usable to business users. The chapter provides an overview of PerformancePoint service (part of SharePoint), and Power View (part of SharePoint or Excel), before providing practical exercises, including:
In each case, a step-by-step walkthrough is provided, with useful instructions and screenshots. This chapter provides a useful practical introduction to using both PerformancePoint and Power View in dashboard design, the examples should prove especially instructive. Chapter 11 Power BI Power BI enables capable business users to perform self-service BI. Power BI consists of various tools, Power View is the main component (discussed in the previous chapter), while this chapter discusses Power Query (self-service ETL tool) and Power Map (3D tool for geospatial data). The chapter provides an overview of self-service ETL using Power Query, considered a light version of SSIS. A step-by-step example shows how to retrieve smartphone data from the web, apply transformations (e.g. adding extra columns), and combine this with other data sources. The chapter continues with a look at Power Map, and provides a step-by-step example that produces visually pleasing 3D graphs of house information (e.g. price, size) in Dallas. This chapter provides a useful introduction to the capabilities of Power Query and Power Map. It should be noted that Power BI components have many more features. Chapter 12 Integrating Reports in Applications This chapter outlines the various ways of embedding SSRS reports or other visualization components into .NET applications (i.e. Web, Windows, Metro). The main components described are ReportViewer and data mining models. In each example, a step-by-step walkthrough is provided, with useful instructions and screenshots. This chapter provides an interesting overview of how SSRS reports or other visualization components can be integrated easily into the various types of .NET applications. Conclusion This book aims to explore Microsoft’s Business Intelligence (BI) tools, and succeeds. The book is generally easy to read, with useful explanations, practical step-by-step walkthroughs, helpful screenshots, inter-chapter references, and website links. There is much more to learn about Microsoft’s BI tools, many of the components discussed here have many other options that have not been discussed, but the book does provide a useful entry point to the most commonly used BI components. The book does require some previous knowledge of databases systems, so it’s not a guide for the complete beginner - contrary to its title. Sometimes, insufficient information is provided before an example walkthrough (i.e. why we do something), however, useful information is provided after the walkthrough. There are a few areas of bad grammar but nothing too onerous. It might have been useful to include some comment on the growing importance of Big Data, and Machine Learning. I do wonder if they might lead to the decline of proprietary data warehouses and BI tools in the near future. If you have some knowledge of database systems (OLAP in particular), this book provides a great entry point to learning Microsoft’s BI tools. Recommended. 
|
||||||
| Last Updated ( Wednesday, 06 April 2016 ) |