Data Mining Assignment - Essay Example

? DATA MINING Credit fraud involves all crimes committed using a credit card during payments or transactions. The crime is based on obtaining unauthorized funds or goods without any payment. Many people especially the old have fallen victims due to fraud. This is because this group of people has irregular checks of their cards and bank accounts. Most of the time these people are assisted by people whom they trust to perform their transactions who later becomes the perpetrators. Generally there are incidences of fraud that have not yet been detected (Borzykowski, 2012 p. 34). With increased cases of fraud, it is a high time for effective measures to be applied to combat these crimes for once and for all. All the merchants, banking systems and the card owners should enroll on application of effective measures. With the application of data mining, one is capable of determining the hot spots which are the target for these crimes. Due to increased technology in businesses, application of computer science Information Technology would help solve this phenomenon (Borzykowski, 2012 p. 34). Efficient fraud detection unveils suspicious behaviors providing alarms to the organization.Cases of fraud experienced in data mining are collected (Tan, 2013, p.345). Metrics for calculating the fraud data are designed and an automated mode of their calculation is developed. Finally the IT expert’s develops a detection model for the fraud. Globally more than 30% of firms have experienced fraud in the year 2009.Retail businesses like supermarket have enrolled in usage of closed-circuit televisions in conjunction with POS data in fraud detection (Tan, 2013, p.345). 2. Introduction Fraud detection can be categorized into statistical techniques and artificial intelligence. Statistical data analysis involves pre-processing of data like detection, validation, error correction; missing and invalid data rectification. One can match algorithms in detection of any abnormality in transactions. Forensic accountants specialize in procurement and analysis of electronic data in detection and rectification of an error. Merchandising agents have started using un-supervised methods like Break Point Analysis and Bolton Hand Use Analysis in detection of credit cards accounts frauds. Peer group analysis is capable of detecting individuals who behave in a different way compared to the previous individuals seen. Break point analysis detects the abnormal transactions in a given account (Robert, John and Gary, 2009, p.543). A three level profiling operates at the account level to detect any form of fraud. Normal profiling and behavioral profiling are applied. Human pattern reorganization and automated data algorithms are linked to create Domain-Specific Interfaces to visually present the accounts holder’s data (Robert et al 2009, p.543). Banks should start using advanced software to detect any miscellaneous transactions. More security measures like pin and ZIP codes should be provided by the customer whenever he or she is conducting any transaction. The software will detect any transaction done at a far distance from the card owner’s geographical location. More details of the card owner like a passport photo should be displayed every time. The photo will enable the merchant or the bank to compare the physical appearance of the customer and determine if he or she is the authorized person. One will be able to determine the common area in which the owner conducts his or her transaction most frequently (Borzykowski, 2012 p. 35). Card owners should also be vigilant by ensuring that they do perform regular checks of their accounts. They should keep their important documents in safe places and besides being cautious on the people whom they have authorized to perform their transactions. The government needs to take serious measures on any person accused of conducting fraud crimes (Borzykowski, 2012 p. 35). 3. Data Mining Data mining is a field of computer science that deals with discovering large data sets using methods like machine learning, statistics, artificial intelligence and database systems. The process extracts data and transforms it into a meaningful form for future use. The process also involves data management, data pre-processing and inference. Data mining functions in the analysis of huge data sets to extract unknown patterns such as unusual records, data groups, and dependencies (Borzykowski, 2012 p. 36). The term data mining is frequently confused to large scale data processing which is not so. It can be generalized to any computer system that supports machine learning, business and artificial intelligences. The basic item is the discovery of a new idea that is based on the use of a computer system. Data mining involves an automatic or semi-automatic analysis of large contents of data, hence the term discovery. Multiple data can be analyzed to obtain accurate predictions using a computer based decision systems (Borzykowski, 2012 p. 36). Data mining uses the previous information to deduce some problems which may occur in the future. Business institutions can use data mining to determine marketing strategies in comparison to their competitors’ techniques for them to win the market. Data mining enables them to get the real time analysis of the situation in the promotion of a new product, increase sales or delete production of a low-value product (Borzykowski, 2012 p. 36). Data mining is a customer relationship management. It bridges the gap between applied statistics and artificial intelligence in relation to the data base management by unveiling the way data is stored and analyzed to the actual algorithms. This allows the method even to be applied in larger data sets. Data fishing, snooping and dredging are terms used when referring to data mining whereby one samples part of a large population data those are too insignificant to deduct some statistical inferences of the data validity (Borzykowski, 2012 p. 36). Data mining is known to compose six classes of tasks mainly anomaly detection, association rule learning, clustering, classification, regression and summarization. In business, the term refers to analysis of all historical business activities and events that are stored in warehouse databases to reveal hidden or unseen trends and patterns. Sophisticated software’s are used to analyzing large amounts of data to discover all the hidden business information. Whenever a loyalty or card is used the merchant can use the information to determine the purchasing trend (Borzykowski, 2012 p. 36). 4. Credit Card Fraud Credit fraud involves all crimes committed using a credit card during payments or transactions. The crime is based on obtaining unauthorized funds or goods without any payment. Many people have fallen victims of credit cards victims due to fraud. In 2008, United States experienced a 21% increase of identity theft according to a report by the United States Federal Trade Commission. These frauds have decreased due to effective fraud protections and prevention systems. However, these mechanisms have been implemented to only control and prevent a twelfth of the frauds leaving a conversion billions dollars lose (Borzykowski, 2012 p. 55). Card fraud occur when the card is physically stolen or interference with the associated data of the account number or any relevant information needed for a transaction. The fraud can be conducted in so many ways involving usage of transaction receipts for future use. Increased technology in usage of credit cards for transactions through the internet has led to many accounts experiencing many frauds. A card holder may notice the loss of his or her card quickly on the contrary of an account which receives billing statements irregularly. Card holders frequently check their accounts to manage them for any fraud (Borzykowski, 2012 p. 55). A stolen card may be used for financial transactions only until when the card holder reports it to the issuer who cancels any transaction made using the card, otherwise many transactions will be conducted before the holder realizes that the card is in the wrong hands. Nowadays the banks have opened a 24 hours communication channels for the victims to report any fraud (Borzykowski, 2012, p. 45). Self service systems like kiosks and filling systems are the common targets for stolen cards since they have no mechanisms to identify these stolen cards. Once a fraud has been committed on a small value good, the merchant takes little attention to prevent it (Borzykowski, 2012 p. 35). In Europe and other continents the card has an EMV chip which requires an input of a four digit pin in any terminal for any successful transaction. Online transactions do not require any pin input since it decodes the information on the magnetic code assuming it belongs to the person conducting the transaction. Zip and postal code details are required to be produced before any service is delivered in US provided one is using a credit card. Loss of the card together with other documents in a wallet may help the thief to get all the requirements needed for a transaction (Borzykowski, 2012 p. 35). With great advancement in computer science and software development, card issuers have developed very sophisticated software’s to combat these frauds. A huge transaction being made on a far distance from the holders’ home postal address may be cancelled or the card withheld by the machine. The holder is then contacted through a phone call or just an email to verify the transaction or the card belongs to him or her in order to get it back (Borzykowski, 2012 p. 35). 5. Countermeasures To combat credit card fraud the following Countermeasures can be applied by the merchants; Requesting the customer to produce additional information, such as a PIN, ZIP code, or Card Security Code . The merchant to conduct geographical location validation details like IP address of the user (customer). PAN truncation whereby full details of the receipt are not displayed. Data security whereby there is incomplete storage of full account details in computer systems. The public can share any relevant information about any known fraudsters to the authorities (Borzykowski, 2012 p. 45). To combat credit card fraud, the following Counter-measures can be applied by the card issuers: Use of fraud detection and prevention software that analyzes patterns of normal and abnormal behaviors of the customer’s transaction. Among the leading software companies that offer solutions for card fraud include SAS, BAE Systems and IBM. Blocking a miscellaneous card transaction till verification of the owner is performed. They should also Contact the cardholder to request for verification details of the card besides applying adequate preventative measures on accounts which had experienced fraud earlier (Borzykowski, 2012 p. 45).Technological Features involves the use of 3D security, strong authentication of the cards and EMV.IBM is a global professional firm and the largest Information Technology (IT) consultancy firm that offers management consultancy, application management services and system consultations (Borzykowski, 2012 p. 45). 6. Business intelligence Business intelligence (BI) is a set of methodologies, architecture, theories and technologies that transform raw data into meaningful information in a business. It can handle a large amount of data at once in the identification and development of business opportunities. Identification of a new opportunity enables a business to implement strategies that will enable it to compete with other institutions hence winning the market. This term is sometimes taken to be synonymous with competitive intelligence since both applies to decision making in a business institution.BI uses technologies, applications and processes to analyze internal structured data within a business firm while competitive intelligence gathers and analyzes data of a particular competitor firm. BI applications use the gathered data in warehouses or data mart. A warehouse is a copy of analytical data that facilitate making of a decision (Borzykowski, 2012 p. 65). Measurement program creates a platform through which business stakeholders use to gauge their goals. Analytic program enables a business to make right decisions and invent new discoveries. Enterprise reporting serves the strategic management of a business through executive information systems, data visualization and OLAP. Knowledge management program can be applied to identify, create and adopt experiences that will be useful in expansion of a business. Collaboration platform program enables a business to share its ideas with its external competitors through electronic data exchange and data sharing. In addition, Business Intelligence provides an alarm function to the end users who are the customers for efficient management of their affairs. A negative feedback will imply that something unusual has happened hence the need of an expert to govern the data and normalize everything. A business will need to use well structured quantitative methods to create stable prioritization of the needs of the organization by applying weighted decision matrix (Borzykowski, 2012 p. 65). To increase the efficiency of a Business Intelligence, it is important for the issuers to conduct a survey on the customers and inquire what their wishes are or the problems they are experiencing a stage called gathering phase. The users can also add an element of competition to the firm making the business be in a better place of competition, full completion and financial security to the users. The IT department should also be contacted for them to give details of unto which limit they can solve the issue. Use of a user-based approach will also enable a quick adoption of the Business Intelligence (Borzykowski, 2012 P. 34). Conclusion Without prior use of integrated IT securities in financial institutions, credit card fraud will always remain a major problem to both the card users, merchants and banking institutions. Use of credit cards is technological advancement in financial institutions. Fraud is t technological advancement too but in the negative side. Therefore, the solution needs to be a technological invention too. Data mining and business intelligence can help to solve them all provided firm foundations are considered (Borzykowski, 2012 p. 75). This will in the long run prevent all fraud crimes committed and reduce the occurrence of any loss. Bibliography Borzykowski, B. 2012. Debit Card Fraud Detection for Dummies. New York: John Wiley and Sons. Robert, N., John, E. and Gary, M. 2009. Handbook of Statistical Analysis and Data Mining Applications. New York: Academic Press. Tan, P. 2013. Introduction to Data Mining. New York: Pearson Education Limited. Read More