Benefits of Data Mining

Data mining is defined as â€Å"a process that uses statistical, mathematical, artificial intelligence, and machine-learning techniques to extract and identify useful information and subsequent knowledge from large databases, including data warehouses† (Turban & Volonino, 2011). The information identified using data mining includes patterns indicating trends, correlations, rules, similarities, and used as predictive analytics. By employing predictive analytics, companies are actually able to understand the behavior of customers. Predictive analytics examines and sorts data to find patterns that highlight customer behavior. The important behavioral patterns are those that indicate what customers have responded to and will respond to in the future. Also, patterns can indicate a customer base that is in jeopardy with the company, customers that are not company-loyal and are easily lost. Predictive analytics of customer behavior can be of great benefit to the business (Turban & Volonino, 2011). Companies are able to build specific marking campaigns and models such as direct mail, online marking, or media marking based on customer preference and are better able to sell their products to a more targeted customer base. Knowing what the customer wants, what they will respond to, and which customer base to focus on takes the guesswork out of marking and product development. Taking the information retrieved and using it correctly will only increase profits (Advantages, 2012). Association discovery using data mining provides a huge benefit to companies. Association discovery is finding correlations or relationships between variables in a large database. For example, in terms of a supermarket, it is finding out that customers who buy onions and potatoes together are also highly likely to buy hamburger meat. These correlations where one set of products predict the buying of another is referred to as associations. Data mining can employ association discovery allowing business to predict buying patterns and allow for more effective operations management and can better pinpoint marketing strategy of coupons and incentives (Association Rule 2012). Web mining is another aspect of data mining. Web mining uses the data collected on the Internet to analyze customer data and gather information beneficial to the company. Any time someone visits a website, uses a search engine, clicks on a link, or makes an electronic transaction data is generated subject to analytics. Companies use web mining to gain customer preference and insight. The information gathered is used to improve websites and create a better user experience for the customers. Web mining can also be used alongside of predictive analytics. For example, on e-commerce sites every transaction is analyzed. When a customer clicks on a product, web mining tools can present a list of products he/she may also be interested in because of other customers with the similar buying interests/habits. This tool can be extremely effective in gaining business intelligence of the buying habits and preferences of customers (Turban & Volonino, 2011). Data mining also employs clustering to find related customer information and to provide valuable information to the company. Clustering gathers information and designates clusters of similar products and objects. In data mining, clustering is usually the first step. It identifies similar information and groups them to be further examined. Customer information and demographics are an example of these clusters. The group characteristics are analyzed against desired outcomes to understand the buying habits of customers and what marketing campaigns will enhance customer response (Ali, Ghani, & Saeed). Reliability of Data Mining The benefits of data have been examined, but it is important to look possible implications as well. Data mining uses algorithms to predict patterns and customer behaviors. Constant measures are needed to make sure the algorithms are working correctly, but the issue of reliability stems a little deeper. Algorithms and data analysis can only be as reliable as the actual data analyzed. Data gathered from different sources can potentially be t or even conflicting. This greatly affects the validity and result of algorithm, especially predictive analysis. It could alter the customer’s historical purchases or demographic information rendering the information useless and even costly. Data mining is a useful tool and should be trusted up to a point. It should not be the only solution. Companies should not only use data mining for marking and operations decisions. The costs of mistaking customer preference and predicting behavior could be catastrophic (Data Mining). Privacy Concerns of Data Mining. One of the major disadvantages of data mining is the privacy concerns associated with the technique. Three major privacy concerns raised by consumers are identity theft, misuse of personal information, and the â€Å"big brother is watching you† feeling (Orwell, 1954). The first concern is identity theft. With the increasing trend of e-commerce and electronic funds, identity theft has been a huge issue. The sheer amount and speed of information processing through data mining has led to a rise in identity theft making this valid concern. The information could easily fall into the hands of anyone (Exforsys Inc, 2006). The second concern is the misuse of personal information. Companies gather information as specific to customer purchases, names, phone numbers, addresses, and other information then store it in a database. Once obtained, copies can be made with little effort. Companies can easily sell this information to other companies. This is the exact concern of consumers. Consumer information can certainly be misused, exploited, or for discrimination making this a valid concern (Advantages, 2012). The last concern addressed in this paper is the total loss of privacy, feeling controlled or watched. The government uses data mining to track patterns of criminal activity have considered using the technique to track the movement of people. Some people feel this goes too far, and not giving the consumer the choice of having his/her information in the database takes away personal freedom. This concern is tied into the misuse of information because what stops companies to selling information to governmental or private agencies with the sole purpose being to control or watch an individual. With the volatile nature of crime, and the increasing use of technology by government agencies, this concern is also valid (Advantages 2012). Measures have been taken to alleviate these concerns. Companies that utilize data mining are required to take certain actions that protect their customer’s privacy. One of these actions is to remove and identity related attributes from each customer record before the data is transferred to analysts. Also banks allow for identity theft protection services to alleviate the concern of financial security. All of these concerns are still important and steps will have to be continuously made and adjusted to protect the security and privacy of personal and financial information (Li & Sarkar, 2006). Real World Examples of Predictive Analytics Predictive analysis and how it is beneficial to companies has been discussed above in theory. To completely understand how predictive analysis is used is to look at real world examples. The first example is how a fast food restaurant used HyperActive Technologies to predict what customers might order. HyperActive Technologies developed a system that allowed cameras to track vehicles pulling into the parking lot and track customers through the entire ordering process. Using predictive analysis of the data gathers from the cameras, the restaurant was able to conclude that at lunchtime; approximately twenty percent of cars entering the parking lot would order at least one cheeseburger. With this information, the cooks were able to get a head start in food production cutting down on wait time for customers and increasing overall productivity (Turban & Volonino, 2011). Another example of a company that uses predictive analysis is that of INRX, the leading provider of traffic information. INRX uses data mining by evaluating real time traffic measuring traffic problems and congestion. This data is collected from road censors, toll tags, traffic incident data, and commercial vehicles equipped with a GPS that continuously report their speed and location. Using predictive analytics, the data is studied to determine traffic patterns at certain locations and times. Drivers now have access to real time traffic information. This information has proven to be extremely effective and useful to drivers allowing them to make better decisions and avoid unnecessary delays (Turban & Volonino, 2011). The flower company, 1-800-FLOWERS. om, has also used data mining techniques, specifically predictive analytics. The company collects and analyses data at all contact points. Data collected includes historical purchases to discover trends, anticipate customer behavior, and meet customer needs and preferences. This technique has proven to be an effective way of increasing the response rate to customers, identifying profitable customers, and establishing customer loyalty. Customer reten tion increased by over fifteen percent after the implementation of predictive analytics solidifying its effectiveness (Turban & Volonino, 2011). As shown through academic research and real world examples, data mining is a real and effective way of predicting customer behavior and buying patterns. Measures need to be taken not only to overcome the stigma that data mining is unsecure and takes away personal freedom, but to make sure individual information is, in fact protected. If these measures are taken, data mining is a win-win for both businesses and consumers. Consumers will feel heard, understood, and taken care of. Businesses can actually focus resources on building that business-to-customer relationship and will be able to give the people what they need. References Advantages and disadvantages of data mining (2012). Retrieved December 9, 2012 from http://www.dataminingtechniques.net/data-mining-tutorial/advantages-and-disadvantages-ofdatamining/ Ali, R., Ghani, U., & Saeed, A. (n.d.) Data clustering and its applications. Retrieved December 5, 2012 from http://members.tripod.com/asim_saeed/paper.htm Data mining: issues. (n.d.) Retrieved December 7, 2012, from http://www.anderson.ucla.edu/faculty/jason.frand/teacher/technologies/palace/ issues.htm Exforsys Inc. (2006). Data mining privacy concerns. Retrieved December 5, 2012 from http://www.exforsys.com/tutorials/data-mining/data-mining-privacy-concerns.html Li, X. & Sarkar, S. (2006) Privacy protection in data mining. Retrieved December 6, 2012 from http://dl.acm.org/citation.cfm?id=1245621 Turban, E., & Volonino, L. (2011). Information technology for management improving strategic and operational performance (8th ed.). New Jersey: John Wiley & Sons, Inc.