Information Mining and Google - Case Study Example

Information Mining and Google By [Name] [Course] [Professor’s Name] [Institution] [Location of the School] [Date] Table of Contents 1.Introduction 3 2.Literature Review 4 2.1Technological Infrastructure in Information Mining 5 3.Discussion 6 3.1Knowledge 6 3.3The Strength of Information Mining 8 3.4The Workings of Information Mining 8 3.5Ethical Issues associated with Data Mining 10 4.Conclusion 10 4.1Recommendation 11 References 12 Information Mining and Google 1. Introduction Information mining refers to the process of finding patterns or correlations amongst dozens of data fields located in relational databases. It is the procedure through which analysts evaluate data from varying perspectives in an endeavor to summarise it into reliable sets of information. The summarizing is done so as to cut costs, enhance revenue, or both. The software used for mining data has been identified as an analytical tool which facilitates the examination and processing of data. It allows for examination of data from varied dimensions as well as its categorization. Eventually, relationships between the pieces of data in question are identified (Battiti & Passerini 2010, 671–672). Even though the term information mining is considerably new, the technology behind it has been in use for several decades. Google has been enabling clients to search the internet for a significant amount of time. Organizations have been using powerful computer systems to sieve large volumes of scanned data, and also to analyse reports relating to market research. Nonetheless, continued research and innovation has resulted into increased computer processing power, statistical software, as well as disk storage. These successes have dramatically increased accuracy and reduced costs in a significant manner (Yu et al 2009, 56). 2. Literature Review Studies have indicated that information mining is mainly constituted of five elements (Segev & Baram-Tsabari 2010, 816). These elements include: The extraction, transformation, and loading of transaction data into the data warehousing systems. The storage and management of data within the multidimensional databases. The timely provision of access to data as per the needs and expectations of the information technology specialists as well as business analysts. The analysis of data through the use of appropriate application software. The presentation of data in formats which are useful to the stakeholders. For example, inform of tables and graphs. According to Zernik (2010), there are different and distinct levels of analysis available today. Indeed, search engines such as Google do also utilise these levels of analysis. This section pinpoints on six of them. The first one is the artificial neural networks. These networks serve as predictive and non-linear models which are learnt through effective training. Their structures, actually, resemble the biological neural networks. The second category is the genetic algorithms. These algorithms are optimization techniques which utilises such processes as mutation, natural selection, and genetic combination. Their design is based on concepts which correspond to natural evolution (Zernik 2010, 86). Zernik considers the most common level of analysis as being the use of decision trees. In this case, tree-shaped entities represent sets of defined decisions. Such decisions facilitates the generation of rules which then aid in the classification of datasets. The commonly applied decision tree methodologies include the Chi Square Automated and Interaction Detection and the Regression and Classification Trees. The two are decision tree methodologies which are utilised during the classification of datasets. Zernik argues that they usually provide sets of rules which can be employed while analyzing unclassified and new datasets. Such strategies then facilitate the prediction of the records which will have a specified outcome (Zernik 2010, 96). The next level involves the use of the nearest neighbor methodology. Through this technique, each record is classified in datasets which are based upon combinations of classes with k records. Some analysts refer to this technique as the k-nearest neighbor methodology. The fifth level is the rule induction. Rule induction involves extracting a set of applicable if-then rules. In order to accomplish this, data has to be arranged in format which depicts its statistical significance. The last level involves data visualization. The visualization involves the interpretation of a set of complex interrelationships which are present in multidimensional pieces of data. In most instances, data relationships are illustrated using graphical tools (Yu et al 2009, 60). 2.1 Technological Infrastructure in Information Mining There are various information mining applications available in the market today. Google is an example of these applications, but there are others which are more focused on transactional data. These applications are designed to work on systems of all sizes including PC, client/server, as well as mainframe platforms. The prices charged during the acquisition of systems ranges from thousands of dollars to about one million dollar depending on their sizes, sophistication, and processing power. According to Segev and Baram-Tsabari (2010), enterprise-wide applications are of various sizes, and the sizes ranges from ten gigabytes to the excess of eleven terabytes. Segev and Baram-Tsabari argue that there are systems with the power to deliver applications in excess of 100 terabytes. The infrastructure is driven by two important technological drivers (Segev & Baram-Tsabari 2010, 815). The drivers include: The size of database. In cases where high volumes of data are being maintained and processes, organizations tend to go for powerful computer systems. The complexity of querying. Highly complex queries necessitate the acquisition of powerful systems. Contemporary information mining applications have the appropriate management technology as well as database storage to handle the most crucial operations. 3. Discussion 3.1 Knowledge While examining the subject of information mining, it proves to be of great significance to make a logical connection between the raw data and knowledge. Knowledge about issues and circumstances is what inspires the preparation of strategic policies. Present day organizations have been accumulating data of extraordinary volumes. These pieces of data are usually located in different databases, and are usually in different formats (Ehrenberg 2011). They may include: Transactional or operational data. These include records of costs, payroll, inventory, sales, as well as accounting. Non-operational data. These may include forecast data, macro-economic data, and industrial sales. Meta data. Meta data is data that elaborates itself. Information refers to relationships, patterns, or associations among pieces of data. For instance, the analysis of transaction data at the point of sale can facilitate the yielding of information which can explain why certain products are selling more than the others. These are the kind of patterns which enable search engines such as Google to deliver as per the queries of the users. Such data can, therefore, enable the management to make changes in the organization’s policy so as to align its vision with the contemporary market forces (Hoffmann 2012, 20). Experienced analysts are able to convert information into knowledge based on historical patterns. This is important because it can facilitate the prediction of future market trends. For instance, a retail supermarket can utilise its summary information to come up with promotional decisions based on the consumer purchasing behaviors. In essence, retailers or manufacturers are able to determine which of their items could sell rapidly after promotional exercises (Segev & Baram-Tsabari 2010, 820). 3.3 The Strength of Information Mining Information mining procedure is being primarily used by organizations which focus strongly on the customers. The nature of such organizations includes communication, marketing, financial, and retail. Information mining enables these organizations to assess the relationship between the internal and external factors. Internal factors include staff skills, product positioning, as well as price. The external ones include economic indicators, customer demographics, and competition (Hoffmann 2012, 19). The assessment of the connection between the two enable the stakeholders determine the impact of their current policies on corporate profits, customer satisfaction, and sales. Eventually, the assessment enables the management to drill-down into the summary information. As such, they are able to view important details of transaction data and, thereby, make appropriate arrangements so as to profit from the available opportunities (Jockers et al 2012, 30). 3.4 The Workings of Information Mining While information technology evolves separate analytical and transaction systems, the process of information mining links the two. Information mining software examines patterns and relationships amongst the stored pieces of transaction data. The examination is based upon user queries which are open-ended in nature. There are several examples of analytical software in the market. They include neural networks, machine learning, and statistical software. In most instances, analysts seek four categories of relationships (Hoffmann 2012, 19). They include: Classes Classes are those stored pieces of data which facilitate the location of data within predetermined groups. There are many examples of classes. For instance, restaurant chains may mine information relating to the clients purchasing behavior in order to determine when the said customers visit, as well as what exactly they order during their visits. Such pieces of information can be utilised in a manner that increases traffic as well as purchases. Clusters Clusters are what facilitate the grouping of data items as per such logical relationships as consumer preferences. Data pieces can, for instance, be mined for the purpose of identifying consumer affinities or market segments. Associations Pieces of data are usually mined with the intention of identifying associations. It is these associations which facilitate the formulation of appropriate business decisions. Sequential Patterns In most organizations, pieces of data are mined for the purpose of anticipating trends, patterns, as well as consumer behavior. For instance, retailers may use data to predict the percentage of bad debts in the following financial year. 3.5 Ethical Issues associated with Data Mining Most people get concerned when organizations withhold information about them indefinitely. Indeed, such withholding means that organizations interfere with the privacy of their clients. This is disadvantageous. The observation means that the concerned managements ought to establish ethical guideposts which would ensure that the information is used for the intended purposes only. Client data do also need to be utilised in a manner that is not contrary to the expectations of the concerned individuals (Jockers et al 2012, 30). In fact, clients ought to be informed of what type of information is retained once they interact with the organization’s staff, and when their credit cards are used. It is unnecessary to retain personal information. The organization should restrict itself to the retention of information which facilitates the manipulation of statistics. This would, actually, encourage the clients to continue their cooperation with the organization (Chen & Fu 2009, 19). 4. Conclusion This report has indicated that information mining technology is endowed with great potential. Information mining enables the analysts to discover information that ordinary reports and queries cannot reveal in an effective manner. This paper has explored several aspects associated with information mining. Indeed, it has indicated that the future of the information mining techniques is bright, considering that there has been significant research and innovation on the subject (Battiti & Passerini 2010, 680). Raw data rarely provide useful clues. It needs processing into more helpful information. The fierce competition in the current business environment means that organizations have to turn raw data into useful insights with remarkable speed (Jockers et al 2012, 29). 4.1 Recommendation The mere storage of data in data warehouses does not benefit an organization. This report does, therefore, recommend that the stakeholders mine and process data in an appropriate manner. Doing this would enable them to enhance their knowledge of markets, customer, and prevailing market forces. The paper has indicated that information mining is a form of knowledge discovery. For there to be effectiveness, process must be accomplished through the use of appropriate information mining software. The software must be the one which facilitates the predicting of future trends and consumer behaviors. This process would, in the end, facilitate the coming up with knowledge-driven and proactive decisions (Ehrenberg 2011). With the information acquired through information mining, ought to focus on utilizing the records of purchases to send several targeted promotions. The promotions ought to be mad based on reliable purchasing history, and within strict ethical guidelines. After mining important demographic data, retailers have to focus on developing promotions and products which appeal several segments of the customer base. Ethics have to be considered so as not to invade the privacy of the target group, or even annoy the potential clients (Chen & Fu 2009, 20). References Battiti, R & Passerini, A 2010, ‘Brain-Computer Evolutionary Multi-Objective Optimization (BC-EMO): a genetic algorithm adapting to the decision maker’. IEEE Transactions on Evolutionary Computation, vol.14, no.15, pp 671–687. Chen, T & Fu, Z 2009, ‘Protection of Privacy on the Web’, in Eyob E, Social Implications of Data Mining and Information Privacy, Virginia State University, USA, pp. 15-32. Ehrenberg, R 2011, ‘Data mining finds new disease links: migraines and hair loss among unexpected connections’, Science News, Science Service, Inc., vol. 180, no. 8, viewed 26 July 2013, . Hoffmann, L 2012, ‘Data Mining Meets City Hall’, Communications of the ACM, vol. 55, no. 6, pp. 19-21. Jockers, M, Sag, M & Schultz, J 2012, ‘Don’t let copyright block data mining’, Nature: International Weekly Journal of Science, vol. 490, no. 7418, pp. 29-30. Segev, E & Baram-Tsabari, A 2010, ‘Seeking science information online: data mining Google to better understand the roles of the media and the education system’, Public Understanding of Science, vol. 21, no.7, pp. 813-829. Yu, C, Zhong, Y, Smith, T, Park, I & Huang, W 2009, ‘Visual data mining of multimedia data for social and behavioural studies’, Information Visualization, vol. 8, no. 1, pp. 56-70. Zernik, J 2010, ‘Data Mining as a Civic Duty-Online Public Prisoners’ Registration Systems’, International Journal on Social Media: Monitoring, Measurement, Mining, nol.1,pp 84– 96. Read More