Chapter 6. How can information systems be used to enhance business intelligence?

Chapter 6. How can information systems be used to enhance business intelligence?

High quality and timely information are very important for the effective management of a company. This can be provided by several different information systems.

What is business intelligence?

Business intelligence is defined as the use of information systems to collect and analyze data and information to improve making good business decisions. The data and information can be both internal and external. Companies use business intelligence to gain or sustain a competitive advantage. Business processes occur at different organization levels and they are highly interconnected.

To monitor and control its business processes a company needs accurate and integrated information. This information is necessary to realize the strategic goals of the organization. Companies that make decisions that can be backed up with verifiable data are called data-driven organizations. Such organizations are measurably more profitable and productive than companies that are not data-driven.

Modern organizations face challenges from external factors, but these factors can also create opportunities. Examples of such external factors are globalization, societal changes and governmental regulations. The business environment is very complex and business intelligence can help organizations to better make decisions in this environment. Business intelligence can help by collecting and analyzing internal and external data more effectively.

If a company is able to effectively manage their data then this will increase the organizational performance of the company. However, many organizations are not able to harness the value of Big data. Big data is data which has high volume, velocity and variety. The trends in social and mobile field have led to an enormous amount of potential data.

The large volume of data enables companies to make decisions that are based on more factors. However, managing an storing such large amount is a challenge for companies. Companies obtain data at a very high rate. Useful data can have several forms:

• Structured data, like transaction data, which can be fitted in a spreadsheet or database
• Semistructured data, sensor data and like clickstreams
• Unstructured data, like video and audio data

For many companies it remains hard to successfully use Big data.

First, organizations did not have the right tools and information to plan their future continuously. Since the environment is nowadays changing very fast it is important that companies can react quickly and this requires new ways of planning. Many companies that are successful use a continuous planning process. With such a process, companies continuously analyze and monitor information and data. This continuously planning can only be executed because of accurate and timely business intelligence.

The most important assets for an organization are data and knowledge since these are crucial for business processes and gaining business intelligence. The database is used to place dynamic and customized information on their Web pages. At many Web sites of companies, customers can view product catalogs and place orders. On the Web site customers can find information on all products, there is a web page for each product.

There are electronic commerce applications which have to process millions of transactions each day. Organizations need to manage their online data effectively so that they are able to understand consumer behavior and to create adequate system performance.

To interact with the data in a database we can use database management systems (DBMs). DBMS is defined as a software application that enables you to store, organize, create and retrieve data from one or more databases. An example of a DBMS is Microsoft Access. Something that we can collect data about, like people, is called an entity. Entities are often referred to as tables, where each column is called an attribute and where each row is called a record.

Most databases under a DBMS consist of multiple tables, which are organized in multiple files. With DBMS you can manage many tables simultaneously. Other advantages of a database approach are improved data consistency, improved data sharing, increased security, improved data quality and accessibility. However, there are also costs like the installation costs and conversion costs.

There are several different database types. A relational database management system (RDBMS) tries to balance efficiency of storage needs, ease of retrieval and other things by storing data in tables that are linked via relationships. Many organizations have used such a system to support their business processes. A management system that has become more popular is NoSQL. These databases can be distributed across multiple machines and can handle more types of data.

A data model is a diagram or map that represents entities and their relationships. In these data models the structure of the data is captured. There are multiple data types and each attribute in the database is of a certain type. Data types enable the DBMS to organize and sort the data, allocate storage space and make calculations.

After the data model is formed, a data dictionary is used to capture the format of the data. A data dictionary is defined as a document that explains multiple pieces of information for each attribute, like name and the type of the data.

Business rules are the policies by which a business runs. These rules are generally included in data dictionaries and help to prevent that illogical or illegal entries enter the database.

The data must be entered into a data base and traditionally this is done by a data entry professional. However, nowadays most data are captured automatically. To enter information, one can use a form. This form has banks in which the user can enter information like the name and date. Such a form shows the user which information to enter.

A report is defined as a compilation of data from the database that is organized and produced in printed format. This can be either on paper or electronic. Data that is stored in the database is often used for reporting. To build interactive records and visualizations to present data, users can use report generators.

A query is used to retrieve data from a database. The structured query language (SQL) is the language that is most used to interface with RDBMSs.

Operational systems are used to run a business in real time and interact with customers. For a company’s internet success, it is crucial to respond quickly to consumers. Therefore, it is necessary to have immediate automated responses when users make requests. This can be done using an online transaction processing (OLTP) system. These systems are designed to deal with multiple concurrent transactions from customers.

The OLTP system gathers, transforms and updates data in the system. It is important that the speed of DMBS for processing transactions is high. Another factor that is important is the choice of which technology is used to process the transactions.

Information systems are systems that are designed to support decision making which is based on stable point-in-time or historical data. It is easy to access and use the data. Information systems are primarily used by managers, customers and business analysts.

Master data is the data that is the most important to the business. Such data includes data about suppliers, customers, employees and inventory. Master data management is important because all the corporate levels of the company need to understand the master data.

Business intelligence applications can access multiple databases or they can use a data warehouse. Data warehouses are used to integrate multiple large databases and other information sources into a single repository. These warehouses consist of both current and historic data and are suitable for direct analysis.

The historical data in the warehouse remains the same but there is also new data from operational systems added each period. Extraction, transformation and loading the data into the warehouse are crucial processes for consolidating data from operational systems.

In the transformation stage the data is cleansed, which is the process of detecting, correcting or removing inaccurate or corrupt data that is retrieved from different systems.

A data warehouse that is limited in scope is defined as a data mart. Each data mart is customized and contains selected data from the warehouse. These data marts are popular among small and medium-sized businesses. Since data marts contain less data than data warehouses, we need less powerful hardware for these which can lead to significant cost differences between marts and warehouses.

What are the components of business intelligence?

To extract information from existing data, companies often use information and knowledge discovery tools. Companies formulate hypotheses and business intelligence tools are used to test for relationships between data.

Information is often represented as a report or query. Some common forms are:

• Scheduled reports
• Drill-down reports (reports with great detail)
• Key-indicator reports
• Ad hoc queries (these are created because there are information needs which are unplanned, and these are generally not saved for later use)
• Exception reports

Online analytical processing (OLAP) is the process of conducting complex, multidimensional analyses of data stored in a database quickly. This process it optimized for retrieval using graphical software tools. With OLAP tools users are able to analyze several dimensions of data that go beyond simple data summaries or calculations.

The head component of the OLAP system is the OLAP server. This server has special functions to analyze data and understands how the data is organized in the system.

With in-memory computing, the data is stored in the main memory of a computer instead of on a relatively slow hard drive. The use of in-memory computing removes the disadvantages associated with writing and reading data. This has become more popular because of the decreased costs of random access memory (RAM).

OLAP systems categorize data as dimensions and measures. The values and numbers a user wants to analyze are called the measures or facts, like the sum of sales. Dimensions are a way to summarize the data, like by region or per time period.

The dimensions are structured like hierarchies. If you use a drill down method then you go from state to country, to city etc. If you go the other way around from region, to country, to continent then we call this roll up.

With an OLAP cube you can analyze data by multiple dimensions. Slicing and dicing is referred to as analyzing the data on subsets of the dimensions.

Data mining enables us to find hidden predictive relationships in the data. Often data mining algorithms are used to find patterns and trends to develop prediction models. Algorithms are step-by-step procedures that are used in a computer program to make calculations or to perform other types of computer based processes.

It can take a long time to run data mining algorithms. Therefore, it is important to reduce the complexity of the data that you want to analyze. This process of reducing complexity is called data reduction.

Association discovery is an application that is often used for data mining. It is a technique that is used to find associations or correlations between sets of items. Sequence discovery is similar to association discovery. It is used to discover associations over time.

Clustering and classification are also applications that can be used for data mining. Clustering is defined as the process of grouping records, which are related, together based on having similar values for attributes. This enables us to structure the data.

When the groups are known beforehand and the records are segmented into these groups then we can use classification. Often, decisions threes are used to classify records.

The tools described can help an organization make better decisions but it only provides them a part of the whole picture. Companies try to reach the entire truth by analyzing unstructured data using:

• Text mining
  Analytical techniques are used to extract information from textual documents
• Web content mining
  Extracting textual information from Web documents.
• Web usage mining
  Used to determine patterns in the usage data of consumers. With clickstream data companies can analyze how customers navigate through a Web page.

For an organization it is useful to analyze textual documents because it enables decision makers to gather competitive intelligence, allows the marketing department to use sentiment analysis and operations management can learn about performance of their products.

There are also business intelligence applications to support human and automated decision making.

Business analytics increase business intelligence by using predictive modeling and statistical analysis to create explanatory models, identify trends and understand data. Business analytics helps us the understand why something is in a certain way and what it will be. It is for instance used to predict how people will react to a certain event.

A special-purpose information system that is made to support the organizational decision making of a particular problem is called a decision support system (DSS). Such systems are often used by employees with a management function and improves human decision-making performance.

With a what-if analysis you can make hypothetical changes to the data and then observe how the changes you made influence the results of the problem.

A decision support system (DSS) consists of input, process and output. With this system models are used to manipulate the data. It is often used for corporate planning, cost analysis, product design, discounted cash flow analysis and marketing.

The science that uses information technologies (hardware, networks and software) to simulate human intelligence is called artificial intelligence (AI). It is used for human intelligence like learning and reasoning or other sensing capabilities.

Conventional computers are not able to adapt to changing circumstances or deal with data that is noisy. AI can be used to enable systems to learn by identifying patterns in massive amounts of data. This is called machine learning and has led to improvements in intelligent systems, language processing, Web searching etc.

Netflix has used machine learning algorithms to improve its movie recommendations. Neural networks are often used when machine learning is applied. These neural networks are networks consist of processing elements that work parallel to complete a task.

Intelligent systems consist of software, sensors and computers embedded in machines and devices. There are many different types of intelligent systems. All these types are based on machine learning.

An expert system is defined as an intelligent system that makes use of reasoning methods, which are based on knowledge about a problem domain, to give an advice like a human expert. These systems try to mimic human expertise. This is done by manipulating knowledge instead of manipulating data.

An ES uses a rule, which is a way to encode knowledge and often has the form if-then. For instance, if the profit is $100 or more, then approve the project. Experts often only have limited information and have therefore developed fuzzy logic. With fuzzy logic experts broaden the capabilities of intelligent systems like ES.

An ES system has inputs, a process and output. The process stage of ES is called inferencing and contains matching facts and rules, determine the sequence with questions and then forming a conclusion. ES is generally used by midlevel managers.

There are programs that provide a service when a certain event occurs. This program runs in the background and is called an intelligent agent or bot. There are different types of agents:

• User agents
• Buyer agents (shopping bots)
• Monitoring and sensing agents
• Data mining agents
• Web crawlers (or Web spiders)
• Destructive agents

Knowledge management is defined as the process to obtain the greatest value from the knowledge assets of the organization. The knowledge assets are routines, principles, formulas, methods, underlying skills etc. They can be explicit or tacit. Explicit knowledge can be documented while tacit knowledge refers to the processes in the mind.

A knowledge management system contains a set of technology-based tools such as communication technologies and retrieval systems. The benefits of these systems are enhanced innovation, improved customer service, enhanced employee retention and improved performance of the organization.

However, there are also challenges to knowledge management. These include focusing too much on technology, forgetting the main goal, dealing with too much knowledge and getting employee buy-in.

An organization can us a map of the contacts of a person to determine if there are connections or links in the organization. This method is called social network analysis and can be used to find groups or people that work together.

Information visualization is another pillar of business intelligence applications. Visualization is defined as displaying complex relationships in the data by using graphical methods. This enables managers to get a quick view of the results and enhances business intelligence.

When making decisions, managers often use digital dashboards to present a summary or key indicators of performance. A digital dashboard shows a quick overview and enables managers to identify items that require immediate attention. These dashboards can use several graphical representations like a bar graph, a map or plot.

If several analysis techniques are combined with interactive visualization to solve complex problems then we speak of visual analytics.

Often geographic information systems are incorporated into digital dashboards. A geographic information system (GIS) is defined as a system designed to create, store, manage and analyze geographical information. These systems are used in many different industries like banking, media, agriculture, retail and insurance. GISs enable organizations to combine demographic, geographic and other data.

Bulletpoint

• High quality and timely information are very important for the effective management of a company. This can be provided by several different information systems.
• Business intelligence is defined as the use of information systems to collect and analyze data and information to improve making good business decisions. The data and information can be both internal and external. Companies use business intelligence to gain or sustain a competitive advantage.
• To monitor and control its business processes a company needs accurate and integrated information. This information is necessary to realize the strategic goals of the organization.
• If a company is able to effectively manage their data then this will increase the organizational performance of the company. However, many organizations are not able to harness the value of Big data. Big data is data which has high volume, velocity and variety.
• The most important assets for an organization are data and knowledge since these are crucial for business processes and gaining business intelligence.
• To interact with the data in a database we can use database management systems (DBMS). DBMS is defined as a software application that enables you to store, organize, create and retrieve data from one or more databases.
• There are several different database types. A relational database management system (RDBMS) tries to balance efficiency of storage needs, ease of retrieval and other things by storing data in tables that are linked via relationships. Many organizations have used such a system to support their business processes. A management system that has become more popular is NoSQL. These databases can be distributed across multiple machines and can handle more types of data.
• For a company’s internet success, it is crucial to respond quickly to consumers. Therefore, it is necessary to have immediate automated responses when users make requests. This can be done using an online transaction processing (OLTP) system. The OLTP system gathers, transforms and updates data in the system.
• Master data is the data that is the most important to the business. Such data includes data about suppliers, customers, employees and inventory.
• To extract information from existing data, companies often use information and knowledge discovery tools. Companies formulate hypotheses and business intelligence tools are used to test for relationships between data.
• Data mining enables us to find hidden predictive relationships in the data. Often data mining algorithms are used to find patterns and trends to develop prediction models. Algorithms are step-by-step procedures that are used in a computer program to make calculations or to perform other types of computer based processes.
• Business analytics increase business intelligence by using predictive modeling and statistical analysis to create explanatory models, identify trends and understand data. Business analytics helps us the understand why something is in a certain way and what it will be. It is for instance used to predict how people will react to a certain event.
• A special-purpose information system that is made to support the organizational decision making of a particular problem is called a decision support system (DSS). Such systems are often used by employees with a management function and improves human decision-making
• A decision support system (DSS) consists of input, process and output. With this system models are used to manipulate the data. It is often used for corporate planning, cost analysis, product design, discounted cash flow analysis and marketing.
• Knowledge management is defined as the process to obtain the greatest value from the knowledge assets of the organization. The knowledge assets are routines, principles, formulas, methods, underlying skills etc. They can be explicit or tacit.
• A knowledge management system contains a set of technology-based tools such as communication technologies and retrieval systems. The benefits of these systems are enhanced innovation, improved customer service, enhanced employee retention and improved performance of the organization.
• However, there are also challenges to knowledge management. These include focusing too much on technology, forgetting the main goal, dealing with too much knowledge and getting employee buy-in.
• Information visualization is another pillar of business intelligence applications. Visualization is defined as displaying complex relationships in the data by using graphical methods. This enables managers to get a quick view of the results and enhances business intelligence.

Tentamentickets

• Know the different systems that are discussed in this chapter.