Data Mining: the Personalization of the Organizations Business Processes - Essay Example

Keeping Data Miners from Overwhelming the Organization Submitted to, Submitted By, of the Submitted on, [September 9th, 2011] WHAT IS DATA MINING? Data, when analyzed from different perspectives so as to bring into useful forms is termed as Data Mining. In other words it can be said that the sorting out of data from its huge repository into a meaningful form for the concerned stakeholders is referred to as Data Mining. This process, often termed as knowledge discovery, is assisted by computing techniques that enable easy extracting out of analytical output that may seem otherwise impossible to be done manually. Data Mining helps organizations to focus on their required data in their data warehouses. The algorithms embedded in order to embed Data Mining Techniques are referred to as Data Mining Algorithms. The Data Mining tools often predict future trends and current behaviors. This enormously assists in the making of knowledgeable decisions. Hidden patterns are often the resultants of Data Mining processes. These patterns may most probably be missed out by humans as they lie out of their expectations. Thus Data Mining does the wonders for an organization that may otherwise seem impossible. Data Mining is usually done using pattern recognition technologies and statistical and mathematical techniques. (Clifton, 2004) THE KEY TO DATA MINING The key to Data Mining, for any respective organization is the Mining of Data with respect to the requirements of its users. This emphasizes upon the fact that the evaluation of raw data would be of substantial value to the concerned organization only if it is produced in line with the requirements of the stakeholder. The Data Miners can work on the data and produce varying trends of results but if they are the results that are already obvious or are beyond the scope of the stakeholder then they are most probably not going to be of any good for the organization that is bound to receive them. Kurt Thearling quotes in one of his articles that, “Data mining, the extraction of hidden predictive information from large databases, is a powerful new technology with great potential to help companies focus on the most important information in their data warehouses.” (Thearling, 1997) So it is for Data Miners to determine the extent of information that is necessary and vital for the conduction of the business processes of an organization. Data Mining is often given the names of Knowledge Discovery from Database and Business Intelligence. Data Mining is in fact the key step in the acquisition of knowledge from a vast data bank. Other processes that contribute to Data Mining for the final discovery of information are often referred to as Pattern Recognition, Data Cleaning, Selection, Integration and Transformation. DATA MINING ISSUES TO BE ADDRESSED: Data Miners have always deployed Data Mining algorithms for mining data. However there are a number of issues that need to be taken of by data miners in order to execute successful data mining operation for the respective organizations. These Concerns can be narrated as follows: SECURITY AND SOCIAL ISSUES: Enabling access of precious data to users and stakeholders alike, pose potential threat to the security of the data itself. The data often collected at initial stages for customer profiling or comparing customer and user data with existing information often involves the comparing of information. These comparisons may prove to be inappropriate for the security of the information itself as chances of their leaking exist to their fullest. The mining of the data itself may result in the emergence of certain facts and figures about personal data of the users that may earlier be unknown and may be lethal to the sanctity of the data involved. USER INTERFACE ISSUES: User Interface Issues are the key to the way Data Miners present the mined information to the users. The intricate details of the Mining processes are not required to be shared with the users. The way information is put up to them is the way an eventual organization’s business processes shape into. As long as the information retrieved from the data mining tools is useful and is understandable by the customer it is good enough and is of substantial use to the organization on the whole. The ease of interpretation of data mining results is dependent upon how well have they been visualized for the customer. The better the visualization, the greater is the idea that the data mining results have ideally been put up with the requirements of the user only. Data exploratory analysis is often eased a lot when the visualization of the mined data is done in accordance with the requirements of the analyst. Effective graphical representation of data may be done by means of many visualization tools. There, however, remains much chance of improving research in the accomplishment of effective visualization tools for large sets of data that can be used to represent mined data. There are a number of issues that need to be addressed in order to improve the visualization of mined data for users. These include ‘screen real estate’, ‘interaction and rendering of information’ etc. The background of a user is also vital in deciding how information needs to be mined in accordance with his requirements. For example, a user with a background of accountancy would need information that he can manipulate while another with a marketing background would want to know data possibilities related to his domain. The nature of the interaction of users with mined data and un-mined data are of a crucial nature. This is because this interaction provides users with a means to manipulate the mined data. This may even lead to the visualization of the mined data by the user itself and also the discovery of knowledge from it at different levels of understanding and different angles. ISSUES OF MINING METHODOLOGY  There are a number of mining approaches among which any one can be chosen. Moreover, there are limitations linked to every methodology. The diversity of data and the nature of the domain in which mining is being performed are also key issues that effect the choice of methodology opted for mining data. PERFORMANCE ISSUES: Data Mining becomes uncertain when performance issues involve large amounts of data. This data may be exponential in nature and may be beyond the analyzing capabilities of normal statistical methods. Mining needs to be done on samples of data in such cases. DATA SOURCE ISSUES: Diversity of Data Types. Due to the immensely large amounts of data and its varying types it is impossible to mine data effectively. The varying the type of data the varying the algorithm required is for data mining. Once all the above mentioned issues have been addressed by the data miners it is essential that the output is in a form that is well suited to the requirements of the organization in which it is deployed. DATA MINERS VS ORGANIZATION An organization’s functioning is dependent upon the mining of data done by Data miners. The mined data, when put at the disposal of the stakeholders/users should be interactive enough so much so that a user may be able to manipulate it in every possible aspect with respect to its requirements and should yet be able to get adequate analytical responses for every possible query generated. The better a set of data has been mined for possible outcomes, the more convenient it becomes for user to adhere it to the personalization of the organization’s business processes. REFERENCES: Beson, A., Smith, S., Thearling, K., (n.d.). An Overview of Data Mining Techniques. Retrieved from http://www.thearling.com/text/dmtechniques/dmtechniques.htm Clifton C. (January 12 2004). Introduction to Data Mining. Print. Thearling, K., (2007). An Introduction to Data Mining. Retrieved from http://www.thearling.com/text/dmwhite/dmwhite.htm Read More