Relation Between Databases And Business Intelligence

Published Date: 02 Nov 2017

​​- ​​​​​​​​​​​​​​​​​Boris Evelson: An Information Workplace Report

Current Business Intelligence Applications

Business Intelligence applications can be generalized into the following 5 categories:

​Multidimensional Analysis

Reporting

Analytics

Collaboration

Measurement

MULTIDIMENSIONAL ANALYSIS:

This refers to the way in which business users can slice and dice their data. Multidimensional views of data allows the user to access and manipulate their data in depth to gather richer information. A well known technology that allows multidimensional analysis is called OLAP - Online Analytical Processing.

​"Olap is a category of software technology that enables analysts, managers and executives to gain insight into data through fast, consistent, interactive access to a wide variety of possible views of information that has been transformed from raw data to reflect the real dimensionality of the enterprise as understood by the user."

​​​​​​​​​​​​​​​​- OLAP Council​

​The OLAP Cube is a good visual representation of what multidimensional analysis is and what slice and dice data actually means. In reality, a cube only has 3 axes, allowing us to input 3 variables to access a specific segment of the cube which will give us the wanted data. In OLAP, a cube may have much more than 3 dimensions making it difficult to visualize, but allowing a much greater insight of a data set.

"A cube is a specialised data store designed to handle multidimensional data and the aggregated numerical data. If you have 30 odd dimensions you cannot imagine this data structure…just take it for granted"

​​​​​​​​​​​​​​​​​​​- Syd Stewart: What is OLAP and an OLAP Cube​​

​​​The following is an example of an OLAP technique called drill down or up:

- ​​​​​​​​​​​​​A list of more OLAP techniques

REPORTING:.

The following was taken from Oracle's website:

"Oracle Reports, a component of Oracle Fusion Middleware is Oracle's award-winning, high-fidelity enterprise reporting tool. It enables businesses to provide instant access to information to all levels within and outside of the organization in a scalable and secure environment. Oracle Reports consists of Oracle Reports Developer... to access any data sources, generate reports in any popular format for web and paper, and to burst and distribute reports to any destination."

- ​​​​​​​​​​​​​Image Source

​ANALYTICS:

"Analytics - is also known as 'business analytics' or 'data analysis' - refers to the software and methods that organizations use to understand data... Organizations use predictive analytics and other kinds of analytics software to gain insights into their financial and

operational performance and from their customer behaviors. With these insights, they can make accurate predictions and better-informed decisions about emerging opportunities, competitive threats and shifts in their markets to increase competitive advantage."

- IBM: What is Analytics​

According to the Business Intelligence Wikipedia page, the following are frequently used methods for analytics:

data mining

process mining

statistical analysis

predictive analytics

predictive modeling

business process modeling

complex event processing

prescriptive analytics

COLLABORATION:

Collaborating through data sharing and electronic data interchange within different areas, inside and outside of a company, creates several advantages. To enable collaboration a company needs to have data marts and a data warehouse.

Examples of collaboration advantages:

Financial consolidation and budgeting​

Keeping data integrity and consistency

Quick access to current and historical data

Enables new type of analysis

MEASUREMENT:

It is important for companies to keep track of their goals, and business intelligence programs allows companies to achieve this by "creating a hierarchy of performance metrics and benchmarking to inform about progress towards business goals."

​​​​​​​​​​​- Wikipedia: Business Intelligence​

Key Performance Indicators is a measurement often used by businesses. An example of a KPI would be the percentage of a business' income that comes from online sales.

Businesss Intelligence and Databases

DATABASES

Databases are systems used to store raw data in a structured manner. By processing this data necessary information can be obtained. This information is useful for any end user (person or organization).​

BUSINESS INTELLIGENCE:

Business Intelligence is a system which takes this information as an input and enables an organization to take intelligent business decisions. BI systems are of many types. One type emphasizes visualization. It allows for a visual representation of the information using graphs and figures to allow making sense of it from business perspective. Another type emphasizes discovering knowledge from the huge amount of data available in the organization. It puts into use techniques like neural networks and genetic algorithms to discover the correlation between different parameters of the data that influence the business.​​​

RELATION BETWEEN DATABASES AND BUSINESS INTELLIGENCE:​

DB Aim

The aims of DB systems are mostly tied down to have data organized in a way that makes it more and more useful. They have a very profound mathematical approach for storing data. Most of the modern and commercially used databases are based on a relational model. The idea of having a mathematical approach is to make retrieval of data for required use easier. It also helps in reducing redundancy of the data in turn allowing the database administrators to make optimal use of the physical storage necessary to store this data.​

​BI Aim:

As opposed to DB, the aim of business intelligence systems is to enable the business to make decisions. In order to do this, the systems need a huge amount of data at their disposal. A decision based on a large amount of data will provide better results than a decision based on a small sample. So their success depends on the reliability, cleanliness and the abundance of data.

​Relation of the aims of the BI and DB systems:

"Data mining (DM) is the process of trawling through data to find previously unknown

relationships among the data that are interesting to the user of the data"

- A knowledge management approach to data mining process ​for business intelligence

(Paper by Hai Wang and Shouhong Wang)

The common link of the DB and BI systems is the DATA itself. Even though the above identified aims of these two systems seem to be contradictory to each other; in reality they complement each other. The reduced redundancy of DB systems, allows them to have clean data which is useful from a BI perspective. The DB systems achieve the reduction in redundancy through normalization. The aim of DB systems to make optimum use of storage space allows them to store a large amount of data in a smaller digital space. This helps a BI systems’ user to make more and more intelligent decisions as the amount of data available at their disposal can be large. This huge amount of data can undergo the regression, neural network or genetic algorithm techniques which learn from the data. Thus it allows knowledge discovery through databases.

The BIs in current times try to get data from different sources to form a data warehouse and then mine this data for intelligent decisions for business. BI systems use OLAP and OLAP cube techniques to understand the different dimensions of the data and enable slicing and dicing of it to provide an insight to some hidden facts. These facts which were earlier not known can now be exploited for better business opportunities.

Thus BI and databases go hand in hand.

Factors affecting Evolution of BI

COMPETITION:

The Success story of Harrah’s Entertainment has played a major role in bringing importance to Business Intelligence. Also the research shows that the entire industry can be divided into 2 categories namely Top performers and Low performers. The differentiating factor between them is that only 30% of the competitors are using Business Intelligence for their main Strategic Decision Management. But even the top performers think that they are not using the techniques to their fullest capacity.

INFRASTRUCTURE:

The Infrastructure foundation for the enterprises attempting to implement and use Business Intelligence - a data warehouse - is a subject oriented, integrated, time-variant and non-volatile collection of data that differs from conventional online transactional processing (OLTP) databases.

ARCHITECTURE:

BI system’s architecture is highly complex owing to the backend systems originating from multiple data sources and to the vast volume of data to be processed.

Upcoming Trends in BI​

ANALYTICS:

Business Intelligence is mainly used for adding value and maintaining market leadership to the existing enterprises. Furthermore, it aids in discovering new rules and forming new business decision. It also supports discovered results, e.g the famous beer and diaper relation, by giving the covariance between the two entities and utilizing the database to confirm this finding.

SOFTWARE AS A SERVICE:

The use of BI is rapidly gaining importance and constatntly evolving; however, it requires an entirely new set of sophisticated hardware and software infrastructure. A plethora of companies have identified this need and are providing BI Software as a Service (SaaS) solutions by leveraging the capabilities of the cloud. An example for the same would be GoodData.

MOBILE BUSINESS INTELLIGENCE:

The need for on the fly business intelligence and data supporting the same is gaining a lot of attention in recent times because of which the leader in database management systems, Oracle, has come up with Mobile Business Intelligence tools to continue to stay as a market leader in both the services.

Future of BI

Let us now look at how DBMS would affect BI in the future. According to Gartner’s latest Hype Cycle for Business Intelligence Report, the primary shift in this regard has been towards Analytical In Memory Database Management Systems (IMDBMS) that focus on addressing analytical needs, leveraging the high speed of in-memory capabilities. Over the past few years, technologies like in-memory column-store DBMSs and in-memory row-store-based technologies have become more realistic due to the rapidly reducing cost (and increased capacity) of system-class memory or RAM, allowing for low-latency loading of a large number of data points and analysis. Examples include the processing of sensor data, such as solar cell sensor data, or real-time pricing for online bidding.

Furthermore, the speed of IMDBMS for analytics has the potential to simplify the Data Warehouse model and reduce maintenance by removing the need for aggregates, summaries and cubes. This will result in lower administration costs and offer greater agility in meeting analytical requirements.

Technology offerings for analytical IMDBMS are new and have had limited adoption overall. SAP has been leading the charge over the past year with the introduction of the SAP Hana platform to the market.

SAP HANA

According to the parent company SAP, SAP HANA is deployable as an on-premise appliance, or in the cloud. It is best suited for performing real-time analytics, and developing and deploying real-time applications. At its core is the SAP HANA database which is fundamentally different than any other database engine in the market today since it is capable of processing both transactional and analytical data. Currently, when transactional DBMS products are used for analytical workloads, they require separation of the data into different databases (OLAP and OLTP). The data has to be extracted from the transactional system (ERP), transformed for reporting, and then loaded into a reporting database (BW). SAP HANA has a hybrid structure for processing transactional workloads and analytical workloads fully in-memory. As transactions are happening, they can be reported live. By consolidating two landscapes (OLAP and OLTP) into a single database, SAP HANA allows for real time decision making to happen.

Roadblocks

So far, the major blocking factors for adoption of the technology have been:

Difficulty in articulating value and finding business justification for adoption.

It is proving a monumental task to convince CxOs that data driven decision making is the way of the future. While many acknowledge its need, few actually walk the walk.

Limited history of success

Even though IMDBMS have been in existence since the mid-90's, they have only recently gained prominence due to cheaper RAM.

Low availability and disaster recovery capabilities.

​Since the backup is esentially on disk, it is prone to all the risks associated with current DB systems including system failure, power loss etc