Unit I

Introduction to BI and Decision Support system

Definition of Business Intelligence
● Wikipedia coins the term BI as “a set of theories, methodologies, processes,
architectures, and technologies that transform raw data into meaningful and
useful information for business purposes.”
● In 1989 Howard Dresner coined the term Business Intelligence (BI) again.
He introduced it as an umbrella term for a set of concepts and methods to
improve business decision making, using systems based on facts.
● Business intelligence systems provide actionable information delivered at the
right time, at the right location, and in the right form to assist decision
makers.
● BI is defined as “an integrated, company-specific, IT-based total approach for
managerial decision support”.
Summarizing the different definitions, BI can be characterized by the following
Features:
● Task of BI: The main task of BI is providing decision support for specific
goals defined in the context of business activities in different domain areas
taking into account the organizational and institutional framework.
● Foundation of BI: BI decision support mainly relies on empirical
information based on data. Besides this empirical background, BI also uses
different type of knowledge and theories for information generation.
● Realization of BI: The decision support has to be realized as a system using
the actual capabilities in information and communication technologies (ICT).
● Delivery of BI: A BI system has to deliver information at the right time to
the right people in an appropriate form.
Brief History of Business Intelligence
● 1960s-1970s: The concept of BI began with the development of Decision
Support Systems (DSS). It focused on providing executives and managers
with the tools and information necessary for making informed decisions.
Technologies like Online Analytical Processing (OLAP) emerged during this
time.
● 1980s-1990s: Executive Information Systems (EIS) became popular in the
1980s, allowing top-level management to access summarized reports and key
performance indicators. In the 1990s, data warehouses gained prominence,
consolidating data from different sources for analysis.
● Early 2000s: The term "Business Intelligence" gained traction as companies
started adopting more sophisticated BI tools. This period saw the rise of
software solutions focusing on data visualization, ad-hoc querying, and
reporting.
● Mid-2000s: Self-service BI tools began emerging, empowering non-technical
users to create reports and analyze data without heavy reliance on IT
departments. Companies like Tableau, QlikView, and Power BI started
gaining popularity.
● 2010s: Big Data became a significant part of BI, as organizations dealt with
massive volumes of data from various sources. Advanced analytics,
predictive modeling, and data mining gained importance to derive deeper
insights.
● Recent Trends: The integration of AI and machine learning into BI tools has
become more prevalent, enabling automation, predictive analytics, and more
accurate forecasting. Additionally, the cloud has become a standard platform
for BI solutions, offering scalability, flexibility, and accessibility.
Architecture & Components of Business Intelligence

A typical business intelligence architecture

● Data sources. In a first stage, it is necessary to gather and integrate the data
stored in the various primary and secondary sources, which are
heterogeneous in origin and type. The sources consist for the most part of
data belonging to operational systems, but may also include unstructured
documents, such as emails and data received from external providers.
● Data warehouses and data marts. Using extraction and transformation
tools known as extract, transform, load (ETL), the data originating from the
different sources are stored in databases intended to support business
intelligence analyses. These databases are usually referred to as data
warehouses and data marts.
● Business intelligence methodologies. Data are finally extracted and used to
feed mathematical models and analysis methodologies intended to support
decision makers. In a business intelligence system, several decision support
applications may be implemented such as multidimensional cube analysis,
exploratory data analysis, time series analysis, inductive learning models for
data mining, optimization models.
The main components of a business intelligence system
● Data exploration. At the third level of the pyramid we find the tools for
performing a passive business intelligence analysis, which consist of query
and reporting systems, as well as statistical methods. These are referred to as
passive methodologies because decision makers are requested to generate
prior hypotheses or define data extraction criteria, and then use the analysis
tools to find answers and confirm their original insight.
● Data mining. The fourth level includes active business intelligence
methodologies, whose purpose is the extraction of information and
knowledge from data.
● Optimization. By moving up one level in the pyramid we find optimization
models that allow us to determine the best solution out of a set of alternative
actions, which is usually fairly extensive and sometimes even infinite.
● Decisions. Finally, the top of the pyramid corresponds to the choice and the
actual adoption of a specific decision, and in some way represents the natural
conclusion of the decision-making process.
Business Intelligence Scenarios
● Business intelligence separated from strategic management: In this case
BI is mainly concerned with the achievement of short-term targets in a
division of an organization, for example, a department of an enterprise or a
clinic in a hospital. Typically, results of the BI application are more or less
standardized reports for a dedicated part of the business.

● BI supports monitoring of strategy performance: Such a BI application is

motivated by overall strategic goals and formulated in accordance with these
goals. Monitoring of the performance is done by defining measurable targets.
A data warehouse allowing a unified view onto the business is usually a
prerequisite for such an application scenario.
● BI feedback on strategy formulation: This application goes one step beyond the
previous strategy and aims at an evaluation of the performance using analytical
methods. In the best case, such an application can be used for the optimization of a
strategy. A typical end-product in this scenario may be a balanced scorecard.
● BI as strategic resource: This strategy uses the information generated by BI not
only for optimization but also as an essential input for the definition of the strategy
at the management level. Typical examples are customer based marketing or
development of standard operation procedures for patient treatment.
● This classification depends on the size of the organization and the scope of the
business under consideration.
● For third and fourth strategies we have to take into account specificities of the
application, how many resources can be attributed to BI, and the availability of
information.
● For large production oriented enterprises, the third option may be a good choice,
and in service-oriented businesses the fourth strategy has yielded many success
stories.
Future & Goals of Business Intelligence
● KPIs as Goals for BI
○ Key performance indicator: A KPI links the activities of the business
to objectives by defining a measurable quantity. KPIs may refer to some
aspects of the performance of the business process or to the business as a
whole. One can distinguish between quantitative indicators presented as
numbers, practical indicators interfacing with processes, directional
indicators showing whether the organization is getting better or not,
actionable indicators for controlling effects of change, or financial
indicators.
○ Influential factors: Attributes that may influence the behavior of the
KPI in any BI perspective.
Typology of Analytical Goals
● Descriptive goals generate a summary description for the instances of the
business process from the different BI perspectives.

○ Reporting: Summarize the instances in such a way that one can use the
information for decisions.
○ Segmentation: Group the instances according to a similarity measure
and find representative instances for these groups.
○ Detect interesting behavior: Identify events during business process
execution that allow the identification of important aspects of the
process.
● Predictive goals predict the behavior of instances of the business process.
Two different kinds of prediction may be distinguished:
○ Regression: Find a function that allows the prediction of the output
(usually a KPI) from a number of input variables (influential factors).
○ Classification: Given a partition for observed instances into disjoint
classes, assign a new instance to one of the classes.
● Understanding goals support stakeholders in understanding their business
processes. Two main goals can be formulated:
○ Process identification: Identify the rules that determine the
relationships between the events of the process.
○ Process analysis: Investigate the performance of the instances with
respect to their conformance with a defined business process.
Data, Information & Knowledge
● Data:
○ Generally, data represent a structured codification of single primary
entities, as well as of transactions involving two or more primary
entities. For example, for a retailer data refer to primary entities such as
customers, points of sale and items, while sales receipts represent the
commercial transactions.
○ Data refers to raw, unprocessed facts, figures, symbols, or observations.
○ Example:
■ A collection of numbers: 3, 7, 9, 12
■ A list of temperatures recorded each hour: 68°F, 72°F, 70°F, 75°F
■ A spreadsheet with sales figures for each month: January - $10,000,
February - $12,000, March - $15,000
● Information:
○ Information is the outcome of extraction and processing activities
carried out on data, and it appears meaningful for those who receive it in
a specific domain.
○ Information is processed, organized, or structured data that provides
context, relevance, and meaning.
○ Example:
■ Converting a list of temperatures and associating them with specific
dates, times, and locations: "Temperatures in New York City on
January 1st: 68°F, January 2nd: 72°F, January 3rd: 70°F, January
4th: 75°F"
■ Analyzing the sales figures to derive insights: "Sales increased by
20% between January and February, followed by another 25%
increase in March."
● Knowledge:
○ Information is transformed into knowledge when it is used to make
decisions and develop the corresponding actions.
○ Knowledge is consisting of information put to work into a specific
domain, enhanced by the experience and competence of decision makers
in tackling and solving complex problems.
○ It's the application of information that leads to a deeper understanding or
problem-solving.
○ Example:
■ Understanding weather patterns based on historical temperature data
to predict seasonal trends and plan farming activities or outdoor
events.
■ Using sales data trends to forecast future market demands and
strategically plan inventory or marketing campaigns.
Business Intelligence Tasks & Analysis Formats
● Data Task
○ The data task is a prerequisite for all BI activities. The main goal is
organization of available information about the business and its environment.
○ Typically, the information are data about the structural properties of the
enterprise and the registered customers, the transactional data from business
process instances, the data describing production activities, or traces of
activities in social networks.
○ These data are collected under different data-capturing regimes and stored in
different data sources using multifarious structures ranging from data with
diverse temporal and spatial granularity up to semi structured text data.
○ The data task relies on data modeling techniques encompassing different data
models like ER models, UML, or semistructured data models, including
methods on how to apply the models, and an IT infrastructure for data
provisioning.
● Business and Data Understanding Task
○ Define Business Objectives:
■ Clearly articulate the business goals and objectives that the
organization aims to achieve.
■ Understand how data can contribute to these objectives and drive
value.
○ Identify Stakeholders:
■ Identify and engage with key stakeholders who have a vested
interest in the project's outcomes.
■ Understand their needs, expectations, and how they will use the
insights derived from the data.
○ Data Collection:
■ Identify and gather relevant data sources that are necessary to
achieve the defined business objectives.
■ Understand the quality, format, and structure of the available data.
● Data Exploration
○ Conduct preliminary exploration of the data to understand its
characteristics, patterns, and potential issues.
○ Identify missing values, outliers, and other data anomalies.
● Define Data Variables:
○ Clearly define the variables and metrics that will be used in the analysis.
○ Ensure a common understanding of the meaning and calculation of each
variable.
● Data Governance:
○ Establish data governance policies to ensure data quality, integrity, and
security.
○ Define roles and responsibilities for data management within the
organization.
● Data Privacy and Compliance: Ensure compliance with relevant data privacy
regulations (such as GDPR, HIPAA) and internal policies.
● Data Integration:
○ If dealing with multiple data sources, understand how they will be
integrated to provide a comprehensive view.
○ Address issues related to data compatibility and consistency.
● Data Documentation: Create documentation that provides a comprehensive
understanding of the data, including metadata, definitions, and any
transformations applied.
● Business Rules and Assumptions:
○ Document any business rules or assumptions that are relevant to the data
analysis.
○ Ensure that all team members are aware of and agree upon these rules.
● Risk Assessment: Identify potential risks and challenges related to data
quality, availability, or any other factors that may impact the success of the
project.
● Data Understanding Report: Summarize the findings of the data exploration
and understanding in a comprehensive report.
● Modeling Task
○ The modeling task aims at setting up an analytical business model, i.e., a
formal model that allows precise answers for the analytical goals.

Fig: Overview on modeling activities

● Figure illustrates the interrelations between goals, perspectives, views, and analytical
models.
● The circles in the center of the hexagon represent the different BI perspectives. The
intersection of the circles illustrate that we often have to cope with analytical goals that
need multiple BI perspectives.
● The inner labels at the sides of the hexagon describe the BI views on the perspectives, for
example, taking an event view to analyze the production perspective or a combination of
production and organizational perspective. For example, one may take the event view to
analyze the production perspective or a combination of production and organizational
perspective.
● Above the hexagon denotes the BI goals, i.e., understanding goals, descriptive goals, or
predictive goals.
● The lower part of figure introduces the formal structures used for transforming BI goals
into properties of the model
● The modeling task requires specific data preparation techniques. For example, in the case
of analysis goals referring to text data, different techniques can be used for transforming
such unstructured data to structured data which allow the application of algorithms. Such
transformations will be considered in connection with the analysis methods.
● Analysis Task

Overview on mining algorithms along the BI perspectives

● Having defined a model, one needs algorithms to compute a solution for the
analytical goal within the model. In BI, these algorithms are usually denoted
by the term mining, stressing that we are searching for a solution concerning
a frequently not very well-defined problem.
● The term data mining is originally defined as the analysis step in the process
of the knowledge discovery in databases which has a more exploratory
nature.
● Different types of mining have been proposed; Fig. shows an overview on
these types in connection with the different BI perspectives.
● This includes mining algorithms that frequently occur in connection with
overlapping BI perspectives. At the intersection of the perspectives
production and customer, for example, decision mining has been suggested in
the literature.
● Besides the BI perspective, the choice of the algorithms depends on the view
and the envisaged analytical goal.
● Evaluation and Reporting Task
○ The evaluation and reporting task has to view the analysis results from
two different perspectives.
■ The first one is evaluation of the results in context of the analytical
goal
■ The second one is evaluation from a global business perspective, i.e.
understanding the results of the analysis in the context of the
business.
○ The main goals are the interpretation of the results in reference to
domain knowledge and coming to a decision of how to proceed further.
○ The evaluation task employs reporting techniques that are similar to data
description and visualization techniques.
○ Depending on the intended audience of the report, different types of
reporting can be distinguished.
● Analysis Formats
○ All business activities as a process, we can look at BI activities as a process
and define a structure for organizing the different tasks. Such structures are
subsumed under the term analysis format using ideas from life-cycle models
for software development or from knowledge discovery in databases.
○ There is a basic distinction between cyclical and linear formats. We think that
cyclical formats are more useful in BI, because, in practice, covering the
different perspectives often requires a sequence of models and a combined
evaluation of the results.
○ A number of analysis formats for the knowledge discovery process have been
proposed as data mining formats. One can distinguish between
academic-oriented efforts and application-oriented approaches like CRISP.
○ Other formats are SEMMA or KDD.
○ CRISP mainly focuses on the cross-sectional view on the business process but
adaptations to the state view are possible. Our definition of tasks is closely
oriented towards CRISP.
○ The main difference is that we formulate a closer connection between business
understanding and data understanding.
● Analysis Formats
○ Recently, the L* format has been proposed as a method for business
process analysis and mining, which emphasizes explicitly the idea of the
application of different models and analysis techniques.
○ The L* method starts with the event view of a business process, mainly
seen from the production perspective. After planning and justifying,
which corresponds to business and data understanding, the data (in an
event log format) is extracted and a process model is formulated and
analyzed.
○ Combining the ideas of CRISP and L, we propose the iMine analysis
format that supports the integrated application of data and process
mining.
The iMine method
● Objects: Rectangles with full borders represent the objects that are analyzed
or produced during the BI process along the three goal perspectives.
● Tasks: Ellipses represent the BI tasks that have to be conducted by the BI
analyst
● Techniques: Rectangles with dashed border represent techniques for the
different BI tasks. Techniques consist of repositories for procedures, methods
for using the procedures, and tools for implementation.
● Information flow: Solid arrows represent the flow of information.
● Feedback: Dashed arrows refer to possible feedback loops.
● Analysis goals: Analysis goals are depicted as rectangles with rounded
corners.
Decision Support System
 Definition of Decision Support System.

● a decision support system has been defined as an interactive computer system

helping decision makers to combine data and models to solve semi-structured
and unstructured problems.
● This definition entails the three main elements of a DSS shown in Figure: a
database, a repository of mathematical models and a module for handling the
dialogue between the system and the users.
Features of a DSS
● Effectiveness: Decision support systems should help knowledge workers to
reach more effective decisions. In this respect they are a fundamental component
of business intelligence architectures.
● Mathematical models: In order to achieve more effective decisions, a DSS makes
use of mathematical models, borrowed from disciplines such as operations
research and statistics, which are applied to the data contained in data
warehouses and data marts.
● Integration in the decision-making process: A DSS should provide help for
different kinds of knowledge workers, within the same application domain,
particularly in respect of semi-structured and unstructured decision processes,
both of an individual and a collective nature.
● Organizational role: In many situations the users of a DSS operate at different
hierarchical levels within an enterprise, and a DSS tends to encourage
communication between the various parts of an organization.
● Flexibility: A DSS must be flexible and adaptable in order to incorporate the
changes required to reflect modifications in the environment or in the
decision-making process.

Extended structure of a decision support system:

The structure of the DSS shown in Figure is extended that include following new
components:
● Data management:
○ The data management module includes a database designed to contain the
data required by the decision-making processes to which the DSS is
addressed. In most applications the database is a data mart.
○ The data management module of a DSS is usually connected with a data
warehouse which represents the main repository of the data available to
develop a business intelligence analysis.
● Model management:
○ The model management module provides end users with a collection of
mathematical models derived from operations research, statistics and
financial analysis.
○ These are usually relatively simple models that allow analytical
investigations to be carried out that are very helpful during the
decision-making process.
● Model management:
○ These include simple optimization models, financial and actuarial analysis
models and statistical functionalities.
○ the model management module helps the activities of knowledge workers by
means of high-level languages for the development of ad hoc models.
● Interactions:
○ In most applications, knowledge workers use a DSS interactively to carry
out their analyses.
○ The module responsible for these interactions is expected to receive input
data from users in the easiest and most intuitive way, usually through the
graphic interface of a web browser, and then to return the extracted
information and the knowledge generated by the system in an appropriate
graphical form.
● Knowledge management: The knowledge management module is also
interconnected with the company knowledge management integrated system. It
allows decision makers to draw on the various forms of collective knowledge,
usually unstructured, that represents the corporate culture.
Major potential advantages deriving from the adoption of a DSS:

● an increase in the number of alternatives or options considered.

● an increase in the number of effective decisions devised.
● a greater awareness and a deeper understanding of the domain analyzed and the
problems investigated.
● the possibility of executing scenario and what-if analyses by varying the
hypotheses and parameters of the mathematical models.
● an improved ability to react promptly to unexpected events and unforeseen
situations.
● a value-added exploitation of the available data.
● an improved communication and coordination among the individuals and the
organizational departments.
● more effective development of teamwork.
● a greater reliability of the control mechanisms, due to the increased intelligibility
of the decision process.
Information Systems Support for Decision Making
Information systems play a crucial role in supporting decision-making processes within
organizations. These systems help gather, organize, analyze, and present data to aid
decision-makers in making informed and effective choices. Here are key aspects of
information systems support for decision making:
● Data Collection and Integration: Information systems collect data from various sources,
both internal and external to the organization.
● Data Storage and Management:
○ Information systems store and manage data efficiently, ensuring its accuracy,
security, and accessibility.
○ Databases, data warehouses, and other storage systems are used to organize and
retrieve information.
● Data Analysis and Reporting:
○ Decision support systems (DSS) and business intelligence tools analyze data to
identify trends, patterns, and insights.
○ Reports, dashboards, and visualizations present information in a format that is easy
for decision-makers to understand.
● Modeling and Simulation:
○ Information systems can create models and simulations to test different
scenarios and assess potential outcomes of decisions.
○ This helps decision-makers evaluate the potential impact of their choices
before implementation.
● Decision Support Systems (DSS):
○ DSS are specialized information systems designed to assist decision-makers
in making complex decisions.
○ They provide interactive tools and models, aiding in analyzing information
and exploring various decision alternatives.
● Artificial Intelligence and Machine Learning:
○ Advanced analytics, AI, and machine learning algorithms can uncover
hidden patterns and provide predictive insights.
○ These technologies can enhance decision-making by automating routine
decisions or providing recommendations.
● Collaboration and Communication:
○ Information systems facilitate communication and collaboration among
decision-makers and stakeholders.
○ Collaboration tools, such as shared documents and real-time messaging,
enable efficient decision-making processes.
● Information systems support decision-making by providing timely, accurate, and
relevant information to users. They enable organizations to make informed
choices, improve efficiency, and gain a competitive advantage in today's dynamic
business environment.
Simon’s Decision Making Process
● Herbert A. Simon, a Nobel laureate in economics, is well-known for his
contributions to the field of decision-making.
● Simon proposed a model of decision-making known as the "Simon's
Decision-Making Process" or the "Simon Model," which outlines the steps
individuals go through when making decisions.
● The model consists of three main stages:
○ Intelligence Phase:
■ In this initial stage, the decision-maker identifies and defines the
problem or decision to be made.
■ They gather information, explore alternatives, and assess the available
resources.
■ The focus is on understanding the nature of the decision problem and
the factors influencing it.
● Design Phase:
○ Once the problem is understood, the decision-maker develops and evaluates
potential solutions or courses of action.
○ They create a model or mental representation of the problem and possible
solutions.
○ The goal is to generate and analyze various alternatives, considering the
potential outcomes and consequences of each.
● Choice Phase:
○ In the final stage, the decision-maker selects the best course of action from
the available alternatives.
○ The choice is based on the evaluation of the alternatives in terms of
preferences, values, and objectives.
○ The decision-maker commits to a specific action and implements the chosen
solution.
The Decision Support System-Business Intelligence Connection
● Decision Support Systems (DSS) and Business Intelligence (BI) are closely related
concepts that work together to enhance decision-making processes within an
organization. While they are distinct, they often overlap, and their integration can
create a powerful framework for informed decision-making.
● Here's how the connection between Decision Support Systems and Business
Intelligence is typically understood:
○ Definition:
■ Decision Support Systems (DSS): DSS is a broader concept that refers to
computer-based systems designed to assist decision-makers in solving
complex problems. DSS provides interactive tools and models to support
decision-making at various levels in an organization.
■ Business Intelligence (BI): BI involves the use of technologies, processes,
and tools to collect, analyze, and present business data. The goal of BI is
to provide actionable insights and support data-driven decision-making.
● Data Integration:
Both DSS and BI rely on integrating data from various sources within and
outside the organization. This can include internal databases, external market
data, and other relevant sources.
● Data Analysis and Reporting:
BI tools are often a key component of DSS, providing the capability to
analyze and visualize data. BI supports the intelligence phase of
decision-making, helping users understand patterns and trends in the data.
● Decision-Making Support:
DSS includes BI as a subset within its toolkit. BI tools contribute to the
design and choice phases of decision-making by providing decision-makers
with relevant information and insights.
● Information Presentation:
BI tools help present information in a format that is easy to understand, such
as dashboards, reports, and visualizations. This is crucial for decision-makers
to quickly grasp the implications of different courses of action.
● Ad Hoc Queries:
Both DSS and BI systems support ad hoc querying, allowing users to
interactively explore data and obtain specific information needed for
decision-making.
● Predictive Analytics:
BI tools, especially those with advanced analytics capabilities, contribute to
the predictive modeling aspects of DSS. They can help forecast future trends
and outcomes based on historical data.
● Continuous Improvement:
The feedback loop between DSS and BI is essential for continuous
improvement. Decision-makers can use insights from BI tools to refine and
optimize their decision models within the DSS framework.
● User Empowerment:
The combination of DSS and BI empowers users across the organization by
providing them with tools to access, analyze, and interpret data. This
democratization of data supports a culture of data-driven decision-making.
● Business Intelligence is a crucial component of Decision Support Systems,
contributing to the intelligence, design, and choice phases of
decision-making.
● The integration of BI within DSS enhances the capabilities of
decision-makers by providing them with timely and relevant information for
more effective and informed decision-making.