International Journal of Future Computer and Communication, Vol. 2, No. 1, February 2013 Relevance of Data Mining in Digital Library R. N. Mishra and Aishwarya Mishra Abstract—Data mining involves significant process of Data Mining concept has been delimitated in multiple identifying the extraction of hidden predictive information ways by different organizations, scientists etc. Wikipedia, from vast array of databases and it is an authoritative new has visualized data mining a non-trivial extraction of technology with potentiality to facilitate the Libraries and implicit and potentially useful information from data and Information Centers to focus on the most important the science of extracting useful information from large data information in their data warehouses. It is a viable tool to predict future trends and behaviors in the field of library and sets or databases (http://en.wikipedia. org/wiki/Data information service for deducing proactive, knowledge-driven mining). Marketing Dictionary defines data mining as a decisions. Mechanized, prospective analyses of data mining process for extraction of customer information from a vast move beyond the analyses of past events provided by gamut of databases with the help of the feasible software to retrospective tools typical of decision support systems. Data isolate and identify previously unknown patterns or trends. Mining, Process of Data Mining, Knowledge Discovery in Multiple techniques with the help of technologies are Databases, DBMS, Data Mining Techniques etc. etc. have been discussed in this paper. employed in data mining for extraction of data from the heave of databases. Intelligence Encyclopedia has, however, Index Terms—Data mining, KDD, artificial neural Networks, defined data mining as a statistical analysis technique to sequential pattern, modeling, DMT. retrieve useful data to ascertain trends or patterns (www.answers.com/topic/data-mining). Data mining involves sorting through large amounts of I. DATA MINING THE NOTION data and picking out relevant information and though it is Data Mining, defined in assorted ways with different pertinent for the business organizations and financial implications by the experts, computer professionals and analysts still then, Libraries and Information Centers can be scientists relate to a heave of conglomerated data in excluded under its purview which can be increasingly used multiple areas and it also can be matched with the term such in the sciences to extract information from the enormous as knowledge discovery. This infect involves a process of data sets generated by modern experimental and analyzing data from different perspectives to bring about observational methods. Mention may be made that, Library user centric information that can be employed to increase and Information Service being one of the outstanding revenue, costs, or both. Sporadic attempts have been done service field can well be accommodated in the arena of data by the computer programmers to design data mining mining as a good quantum of data are prevalent in the software where a number of analytical tools have been Libraries and Information Centers especially in digital designed for analyzing data. It allows the users to analyze environment. data not only in multiple dimensions and angles but also its Data Mining can be expressed in other multiple angles categorization, and summarization of the relationships. (Pujari; 2001; p.46) in library perspectives. It can be Technically, data mining is the process of finding employed with library activities through; correlations or patterns among dozens of fields in large Using assortment of techniques to identify nuggets of relational databases (www.anderson.ucla.edu). Data Mining information or decision-making knowledge in the database is a terminology which refers practically to Data Extraction and extracting these in such a way that they can be made from a heave of data available in electronic form. Evolution use in other areas such as, decision support, prediction, of data mining can be traced back when the business data forecasting and estimation. Discovering relation which were first stored in computers and technologies were connect variables in a database which can be interpreted for generated to allow users to navigate through the data in real decision support system time. Data mining takes this evolutionary process beyond Attaining by using pattern recognition techniques as well retrospective data access and navigation, to prospective and as statistical and mathematical technique for meaningful, proactive information delivery and this evolutionary process new correlation patterns and trends by shifting through large is due to the support of three technologies such as, i) amount of data stored in repositories s. In such a situation, massive data collection, ii) high performance computing the library and information center requires help from the and iii) data mining algorithms (Pujari; 2001;p.44) subject experts and other outsourcing. Manuscript received April 25, 2012; revised July 10, 2012. II. DATA MINING- PHASES OF DEVELOPMENT R N Mishra is with Dept. of Lib. and Inf. Sciencce, Mizoram University, Aizawl, India (e-mail: [email protected]), Evolution of data mining has gone through different Aishwarya Mishra is with Computer Science, SUIIT, Sambalpur phases of development. Data mining initially emerged from University, Orissa, India (email: [email protected]), DOI: 10.7763/IJFCC.2012.V2.110 10 International Journal of Future Computer and Communication, Vol. 2, No. 1, February 2013 the business arena where the data were stored in computers To discuss in nut-shell the historical perspectives of data and with the help of relevant technologies users tried to mining, Wal-Mart named after Sam Walton of USA, one of navigate in real time. Massive Data Collection, High the established corporation known as American Public Performance Computing and Data Mining algorithms are Corporation was first to use the Data Mining Technology primarily the associated phenomena which on maturity for transaction of business with relation to grocery and coupled with high performance of relational database consumable where the technique was principally associated engines and broad data integrations precipitated to the with many national and international agencies with a strong employment of data mining technology. Data mining that consumer focus such as, retail, financial, communication identifies trends within data go beyond simple analysis, is a and marketing. This also enabled the agencies to determine component of wider process known as ‘Knowledge the relationships among internal factors such as products, Discovery’ from the databases which involves scientists services, staffs and the external factors such as competition from multiple arenas of disciplines, mathematicians, etc. computer scientists and statiscians including the persons While information technology is employed to separate engaged in machine learning, artificial intelligence, transaction and analytical systems, data mining links information retrieval and pattern recognition (Pujari; 2001; between the components. Software designed and applied in p.2) . Through the use of sophisticated algorithms, users data mining analyzes various relationships and patterns. In have the ability to identify key attributes of target the present ICT era market is flooded with multiple types of opportunities. analytical softwares which can be utilized for data analysis, Knowledge discovery provides explicit information that statistical inferences, machine learning and neural networks. has a readable form and can be understood by a user. It may not be out of place to mention that, the neural Forecasting, or predictive modeling provides predictions of networks are based on the concept of Artificial Neural future events and may be transparent and readable in some Network (ANN) which is associated with information approaches (e.g. rule based systems) and opaque in others processing paradigm inspired by the way of the biological such as neural networks. Moreover, some data mining nervous systems. Key elements involved in this paradigm is systems such as neural networks are inherently geared the novel structure of the information processing system towards prediction and pattern recognition, rather than which is composed of a large number of highly knowledge discovery. interconnected processing elements (neurons) to solve Towards the end of 1980s machine learning methods for specific problems (http://www.doc.ic.ac. searching were started as a means of beyond the fields of uk/~nd/surprise_96/journal/vol4/cs11/report.html). computing and artificial intelligence, which were employed Generally four types of relationships can be sought in the in database marketing applications where the available Data Mining which can be applied to the libraries and databases were used for elaborate and specific marketing information centers especially in a digital environment. campaigns. The term Knowledge Discovery in Databases Classes: Library and information centers in the digital (KDD) was first time coined to describe all those methods environment require accumulating pool of data to meet the which aimed to find relations and regularity among the versatile need of the predefined groups. Therefore, observed data (Giudici; 2005; p.2). Subsequent according to the type of clienteles such as, academicians, technological advances in data capture, processing power students, research scholars,