We condemn the war and support Ukraine in its struggle for democratic values.
We also encourage you to join the #StandWithUkraine movement by making a donation at this link
Data Mining and Decision Support Systems

The process of collecting and analyzing data from different sources and summarizing it into information that can be useful is called data mining. In computer science, data mining is also called data or knowledge discovery. It is an interdisciplinary concept of computer science, which filters through large sets of data and identifies relevant information in these data. Data mining is a computer assisted process that digs through bulky information found in texts, the Internet, and databases. It sieves through different data including the transactional or operational data such as cost, accounting, sales, payroll, and inventory. It also sieves through nonoperational data such as macroeconomic data, forecast data, and industry sales. The other data that data mining sieve through is the Meta data, which contains information about the data itself, such as data dictionary definitions and logical database design (Daniel, 2002).

Data mining processes find patterns or correlations among many fields in large related databases. This technique borrows from fields such as databases, statistics, information retrieval, data visualization, and machine learning. It is useful in fields of business and sciences which handle with large data amounts. The summarized information can be used to cut costs, increase revenue or to market commodities in business (Marek & Roger, 2010).

Data mining yields useful information and knowledge in terms of patterns and trends. Patterns, relationships or associations obtained from bulky data yield significant information. E.g. an analysis of the sales transactions data in a retail point can yield comprehensive information showing the time different products are bought, and which products have the highest demand. Data mining also yields information that reveals historical patterns, and also future trends. E.g. a supermarket can sieve through their bulky data and identify relevant consumer buying behaviours. These behaviours can be utilised, in promotions, to set new consumer buying trends or enhance the correct, old consumer buying trends in the future (Claire, 1997).

The final aim of data mining is to extract valuable knowledge and information, from already existing data, and transform this information into a structure that can be understood by human beings. Data mining also aims at discovering new patterns, which can be useful to their users. The acquired information, in the form of new trends and patterns, serves the purposes of prediction and description. Description focuses on finding and interpreting patterns to users in an understandable and interpretable manner. Prediction, on the other hand, involves identifying fields or variables in bulk data and using them to predict behaviours of some entities or future values (Daniel, 2002).

The technique of data mining derives its name from the similarities that it has with the process of mining minerals from mountains. In mineral mining, vital and valuable minerals are searched from large masses of soil or mountains. Similarly, important information, in the form of data, is obtained from bulk data. Both of these processes, data mining and mineral mining, also require many materials and intelligent probing to discover where the valuable things reside.

Data mining is supported by three technologies namely data mining algorithms, massive data collection, and powerful multiprocessor computers. The developments of these technologies have been in place for many years, in areas of research such as machine learning, statistics, and artificial intelligence. The maturity of these techniques with broad data integration efforts and high performance relational database engines make these technologies applicable in the current environments of data warehouses (Finlay, 1994).

Data mining works through the application of models in a technique called modelling. Models are mathematical relationships or sets of examples. Models of known situations are made and then applied in areas where the answer is unknown. An example of a model can be seen in the telecommunications company, where the marketing director wants to focus the sales and marketing efforts on the population most likely to become significant users of text messages. Though the marketing director understands the customers, it is not possible to identify the common characteristics of the best customers because of the numerous variables in place. However, it is possible to identify the customers who mostly use the text messages using data mining techniques. The manager will use data mining techniques, such as neural networks, to identify customers who use text messages from the existing database of customers, containing information about sex, age, occupation, credit history, zip code, and income. After learning who the best customer is, the manager uses that information to make informed decisions on marketing the text messages services (Claire, 1997).

Data mining provides the link between analytical systems and the separate transactions evolved by large-scale information technology. The data mining Softwares analyze patterns and relationships in stored transaction data using the open-ended user queries. These Softwares search for four relationships in data known as clusters, classes, sequential patterns, and associations. In classes, the stored data locate data in predetermined groups. Clusters group data items according to consumer preferences or logical relationships. In sequential patterns, data are mined with the intention of anticipating behaviour trends and patterns (Marek & Roger, 2010).

The process of data mining is made up of five main elements. Data is first extracted, transformed, and loaded onto the data warehouse system. Data is then stored and managed in a multidimensional system of database. The stored data is then provided to information technology professionals and business analysts, who analyze it using application Softwares. Finally, the data, after being analyzed, is presented in useful formats such as tables or graphs. The data provided at this point can be easily interpreted by human beings (Finlay, 1994).

Managers make complex decisions such as the management of industrial processes, investment portfolios or organizational operations. Others are involved in the control of nuclear power plants and command of military units. It is always a difficult call for managers to predict how individual interactions will affect the outcome of issues such as the introduction of a new policy. The power of a human being's brain cannot predict such results in an organization. Empirical evidence shows that decision making and the human, intuitive judgement deteriorates when a situation becomes complex and stressful. At this point, the quality of the human decision making process is far from optimal, and made decisions are substandard (Claire, 1997).

A deliberate attempt to aid the deficiencies of human decision making and judgement has been made by science throughout history. Disciplines such as economics, operations research, and statistics have made efforts to assist human beings to make rational decisions. These methods have now been implemented in the form of computer applications and soft-ware. These programs exist either as stand-alone computer programs or in integrated computer environments where they aid human beings in making complex decisions. Collectively, these programs and the environments they create are known as decision support systems. Decision support systems are also called knowledge based system because of their attempt to make domain knowledge formal and amenable to mechanized reasoning (Daniel, 2002).

Decision support systems are information systems that computer bases. They help organizations or business in activities of decision making. They are collections of integrated hardware and software applications that act as the backbone of the decision making process, in an organization. Data drives the decision support systems. Since decision support systems analyze data, they largely rely on its collection and availability. Decision support systems serve operations, planning levels, and the management. They assist in making decisions, which change rapidly, and those that cannot be specified easily, in advance (Finlay, 1994).

Decision support systems are mostly applied in the military, health care, planning and management in business, and any area where the management faces complex decision making situations. Their usual application is in strategic planning and formulation of tactical decisions. They also have different applications in the health care sector. High level managers use them to make decisions that occur rarely but have potentially high consequences in the future. The time which is taken when making such decisions usually pays off after a long period (Claire, 1997).

A decision support system has three key components. It is made up of a database management system, model-base management system, and dialog generation and management system. The database management system serves as the reservoir for the decision support system. This system stores data relevant to the class of problems that the decision support system has been constructed for, in large quantities. It separates the physical aspects of processing and the database structure with the user. The database management system should have the ability of informing the user about the different data types available, and how to access them. The model-base management system has an analogous role to that of the database management system. The model-base management system transfers data from the database management system into information useful for decision making. The model-base management system also assists the user in building models. The dialog generation and management system provides an interface that aids in model building. It also enhances the ability of a person, using the decision support system, to utilize it and get full benefits from it (Daniel, 2002).

There are three quantitative models that are used by decision support systems, during the process of decision making, to support business professionals and managers. They include sensitivity analysis, what-if analysis, and goal seeking analysis. The sensitivity analysis studies the impact that a change in one or more parts of a model has on other parts of the model. The goal- seeking analysis finds out the inputs that are necessary to achieve a goal such as the desired work force. Finally, the what-if analysis evaluates the impact of a change in an assumption on the proposed solution (Claire, 1997).

Data mining and decision support systems are closely related in their functions. The process of data mining is an indispensable component of the decision support systems. Data keep on accumulating with the current advances in technology and competitiveness in businesses. This data represent activities that affect the business either in one way or another. As the data are accumulated, it becomes difficult for the human mind to look through all the details in any given situation. Therefore, data mining and decision making systems provide a way out of this complex situation. Data mining helps to sieve through these complex data in the databases, the internet, text messages and other sources, and isolate the useful information from it. The isolated information is then used by different people, especially in management, to assist in decision making. Therefore, decision support systems utilise the technology of data mining to help people solve the problem of making complex decisions (Claire, 1997).

Data mining and decision support systems have been successfully applied in different areas including pharmaceutical companies, banks, and other businesses. These systems aid in the automated prediction of behaviours and trends and also the automated discovery of unknown patterns. The process of sieving through large databases to find predictive information is automated. It allows quick and direct answering of questions that, in the past, required extensive hands-on analysis. In marketing, data mining utilizes past promotional mailings data to identify the most likely targets that can maximise returns on investment, in future mailings. Data mining also predicts problems such as bankruptcy and other forms of default. It also identifies segments of a population most likely to have similar responses to a given set of events (Daniel, 2002).

When discovering patterns that were previously unknown, data mining tools pass through databases, and in one step, identify trends that were previously hidden. Data mining analyses retail sales and identifies products that do not seem to be related but are bought together. In business, data mining discovers patterns such as fraudulent credit card transactions and identifies anomalous data that usually represent errors in data entry.

In business, data mining discovers relationships and patterns that help in making informed business decisions. It helps to spot sales trends, predict customer loyalty accurately, and develop strong marketing campaigns. Market segmentation utilizes data mining to identify the similar characteristics of customers who purchase similar products in a company. Customer churn predicts the customers who have a high likelihood of leaving a company to join another competing company. In business, fraud detection also utilizes data mining to identify the transactions most likely to be conducted in a fraudulent manner. Trend analysis also reveals the difference of a customer with time. Other uses of data mining in business include market basket analysis and interactive marketing. Interactive marketing predicts in what every person, who accesses a web site, is most likely to be interested. Market basket analysis unveils which services or products are usually bought together (Claire, 1997).

Human resource departments in businesses also utilise the services of data mining when making crucial decisions. Through the application of data mining technologies, human resource departments can identify the common characteristics of their most successful employees. Data mining can obtain crucial information about age, the universities attended by these employees, marital status at the time of employment, and gender. Through this information, the human resource departments make wise decisions, concerning the employees to recruit in order to enhance their companies’ success. These strategic management applications assist a company to translate its corporate level goals, such as margin share and profit targets, into operational decisions, such as workforce levels and production plans (Finlay, 1994).

Data mining is also used in the catalogue marketing industry. The cataloguers have histories of millions of customer transactions stacked in their databases. With the use of data mining techniques, patterns among customers can be identified. These patterns assist the cataloguers to identify the customers most likely to respond to upcoming mailing campaigns (Claire, 1997).

In healthcare, data mining provides support in decision making. Health care organisations face a lot of pressure to improve the value of their services and reduce the costs of obtaining it. Health care units generate large volumes of data. They have turned to data mining for support. Data mining enhances disease management, resource utilization, and physician practices in health care. Data mining is also applied to identify the appropriate quality management improvement strategies in Healthcare (Jiawei et al, 2012).

The sequence mining method of data mining is utilised in the study of human genetics. It helps to address the goal of understanding mapping relationships between the variability in disease susceptibility and interindividual variations in the human DNA sequence. Therefore, it assists in finding out how the risk of developing common diseases such as cancer is related to changes in an individual’s sequence of DNA. Obtained information improves the techniques used to identify, treat, and prevent such diseases (Marek & Roger, 2010).

In the banking sector, data mining and decision support systems are used to obtain competitive advantages by gauging the customer responses to changes in business rules. Banks search for trends affected by an increase in usage of credit cards, changes in the minimum payment requirements, and other parameters. Through obtained information, they make informed decisions aimed at increasing their profits (Jiawei et al, 2012).

Large consumer packaging companies also utilise data mining and decision support systems to improve their sales processes to retailers. They use information and data on shipments, competitor activity, and consumer panels to understand the reasons for store and brand switching. After this analysis, the manufacturers select strategies of promotion that reach their target customer segments in the best manner. The information obtained from data mining reduces the cost of doing business and improves the value of the customer relationships (Marek & Roger, 2010).

It is evident that data mining and decision making systems are two interconnected disciplines. Data mining electronically identifies significant trends, patterns and behaviours from bulk data. Managers, healthcare sectors, military commanders and other human personnel utilise these trends, patterns and behaviours to make informed decisions in their respective fields. The process of making decisions will become more and more tasking, and difficult as the quantity of available data adds up. Therefore, data mining and decision support systems will continue to be useful in the business world. However, the process of data mining raises some privacy concerns. As the need for more data increases, companies will try to follow daily activities of all people. This may result in a scenario where privacy is invaded through the tracking of personal information such as telephone calls, purchases made via credit cards, web pages visited, and records obtained in school (Jiawei et al, 2012).

Order now

Related essays