The Advantage of Data Mining - Research Paper Example

? Data Mining Data Mining Predictive analytics involves predicting what each is likely to do ahead by utilizingthe information collected about them. Admittedly, data mining becomes useful in a variety of ways when predictive analytics is applied. The first advantage is effective product recommendation. It is possible to identify what are the preferences of each customer from the available statistics. Using this, it is possible to contact each customer when there is something that will interest them. In fact, the early stages of predictive analytics involve product recommendations and behavioral targeting. Another advantage is the possibility of behavior-based advertising. In order to achieve this, the available data is analyzed to predict the areas which interest each customer and the advertisements of that area are presented to the customer. Yet another area where this predictive analytics can be useful is issues like fundraising for nonprofits. In order to do this, often, companies filter the data to identify donation amounts. Similar is the case of insurance pricing and selection. In fact, it is possible to offer individually tailored insurance packages by analyzing the available data. In addition, it is possible for insurance companies to assess insurance risk using the data. Another area where predictive analytics is highly useful is email targeting. In order to do this, companies identify the nature of the emails each customer is likely to respond to and emails are designed accordingly. However, one of the most important areas of utilizing predictive analytics is retention of customers. It is possible to identify customer defection and attrition through predictive analytics. This will help in reaching customers immediately and effectively stopping their going away. Admittedly, associations discovery helps businesses in a number of different ways. Associations discovery involves indentifying the relationship that exists between the sales of different things or services. In simple terms, association is the discovery of various association relationships in a set of items or services. First of all, this helps organizations identify the related items a customer is likely to buy so that the customer is offered all the related items from the same company. This helps save the time and effort of the customer and ensures that the customer is retained. In addition, it ensures that companies are enabled to sell more by keeping all those items closer which are often sold together. Admittedly, various organizations around the world are effectively utilizing this facility. The first example is Wal-Mart. As Khattak, Khan and Lee (2010) point out, Wal-Mart uses basket analysis and clustering in order to smoothen the business transactions; and this helps the company identify the most sold products, identify the customers based on their purchasing capacity, divide the customers based on their arrival time, and identify the items of major trade. Web mining has its own unique advantages either in the form of selling more products or in the form of reduced costs. The web data collected on customers should be categorized and clustered in order to use the same for various purposes ranging from developing marketing strategies, customer relationships, and competitive analysis. It is possible for a company to utilize usage mining or web log data in order to identify a potential customer and reach out to that customer with a tempting offer. The various forms of web mining ranging from structure mining, usage mining and content mining offer considerable amount of marketing intelligence. This results in more personalized strategies from the part of companies, more sales, more satisfied customers and higher customer loyalty and retention. Admittedly, data mining algorithms are likely to err seriously in the real production environment. So, it becomes necessary to check their validity before using them in the real working environment. The first way of validating a particular algorithm involves the use of various measures of statistical validity to verify whether there are errors in the model. In addition, it is necessary to have the algorithm tested using both training and testing sets of data. This helps the accuracy of predictions. Thirdly, it is necessary to have the results reviewed by experts in order to understand whether the available data is of any use in the particular business scenario. There are various privacy issues raised by people regarding data mining. The first problem is that the data obtained can be used in an incorrect manner. To illustrate, an insurance company may deny insurance or a loan company may deny loan to a person based on the information they availed about the person from web. Such denials are not always legitimate and the rights of the individual are violated through the mining. Evidently, it is almost totally impossible to stop this from happening. The only way out is to stop data mining in total as all people and businesses tend to depend on appearance-in real or on web-to make judgments about a person. The issue is that privacy is highly valued in certain societies. For example, as Thuraisingham (n. d.) points out, some cultures do not like divulging information about their salaries while some other cultures do not consider it as an important issue. Thus, saving privacy according to culture is a highly sensitive area to be addressed. The second important problem identified in this area is the secondary use of personal information. It is found that the personal data collected are often used for purposes other than the ones for which the same was collected. As Brankovic and Estivill-Castro (n. d.) point out, an example of the dangers involved is the killing of actress Rebecca Schaeffer. Her killer had collected her address from the Department of Motor Vehicles in California. In fact, it is necessary for phone companies to record the length of phone calls and for banks to keep transaction records. However, when this information is used for a purpose other than billing, it becomes an invasion into the privacy of individuals. Thus, it becomes evident that people are against the use of data for unauthorized purposes. Yet another point to be addressed here is the possibility of stereotypes. Most people feel that companies may offer different services to different people based on their racial or ethnic groups as there is a general tendency to guess confidential properties based on race and gender. Though it is clear that applying different commercial standards for different people shows business efficiency and competitiveness, it can also promote stereotypes. There are two ways that can be used to protect privacy. They can be called ‘query restriction’ and ‘noise addition’. The former involves offering correct answer to some questions and rejecting the questions which might compromise database. It offers highly reliable statistical information but only to limited queries. The second method involves preventing compromise by introducing an error either to the data or to the result of the queries. Recruitment and selection, budgeting, and customer retention are some major examples where businesses have used predictive analysis to gain a competitive advantage. The predictive analysis is a process of evaluating historical and current data to make potential forecasts about the future behavior. Organizations use predictive analysis to identify what types of candidates are likely to turn out to be top performers. This tool is also assistable to shortlist quality candidates using ATS platforms and HRIS systems. Hence, predictive analysis is very effective to save a considerable time of the recruiter early in the recruitment process. HR professionals opine that this method is beneficial to regulate risk and costs by predictively recruiting qualified and potential candidates (Alignmark). Similarly, predictive analysis is very much helpful to improve a company’s budgeting process. Through predictive analysis, the top management can analyze previous financial data and identify major sources of revenues and expenses. This knowledge would assist the regulators to set aside funds for different business needs with respect to their priority. In addition, they can predict the firm’s growth in revenues and hence make provisions for development projects effectively. The retail industry uses this technique more commonly. Finally, customer retention is another example for the business use of predictive analysis. Experts suggest that predictive analysis can be employed to assess customers’ past service usage and spending behavior. Based on this information, marketers can frame effective strategies for promoting customer satisfaction and customer loyalty, which in turn would lead to customer retention. Companies like Amazon use predictive analysis to meet customer retention. References Brankovic, L & Estivill-Castro, V. (n.d.). Privacy issues in knowledge discovery and data mining. Retrieved from http://www.ict.griffith.edu.au/~vlad/teaching/kdd.d/readings.d/aice99.pdf Jaffee, C. (2011). AlignMark Recruiting and Selection Blog, Jan 18. Retrieved fromhttp://www.alignmark.com/blog/bid/48869/Using-Predictive-Analysis-in-Recruiting-and-Selection Khattak, A. M, Khan, A. M & Lee, Y. K. (2010). Analyzing Association Rule Mining and Clustering on Sales Day Data with XLMiner and Weka. International Journal of Database Theory and Application, 3 (1): 13-22. Thuraisingham, B. (n.d.). Data Mining, National Security, Privacy and Civil Liberties. SIGKDD Explorations. 4 (2): 1-5. Retrieved from http://www.universityoftexasatdallascomets.com/~mxk055100/courses/privacy08f_files/thuraisingham_privacy.pdf Read More