1

1. INTRODUCTION As we all know we are living in a modern era i.e., The 21ST century. As the traditional way to approach things is eradicated as the time progressed ahead in the timeline many traditional worked habits also changed. People are more aware of the technologies that are being used right now. Many organizations are shifting towards more reliable and more accurate systems using various modes of computing in different fields. Whenever we consider a library management system, the modules which completes a proper management system are a well-organized solution for a library. A Database which is constantly managed and updated. A user can get full information of the books of the library easily. All the books are virtually accessible by the user without doing it manually. In the traditional library systems numerous fields are being considered from the perspective of the administrator and the user of the system. Various fields define the system i.e., book details, book ID, author name, etc. It is observed that this existing system uses old algorithms for the systems which are outdated and are less efficient. The tedious task of the system is to manage the records which represents which books are particularly used in a specific time period from the entire library. Earlier library systems were partially computerized that means keeping the records of the library books were done manually, the generation of the reports of the entire database was a manual task which is done by the working officials of the library. As the systems of that time are not fast or more advanced enough to calculate or analyze that amount of data in fraction of seconds. Here we are proposing a system where we are analyzing the existing Database of the library and generating a stock analysis of the books which represents which particular books are used extensively in a particular period of time, using the data analytics and the various data mining techniques. 1.1 Motivation Whenever we see a library management system, it usually consists of a modular based design i.e., each outcome of a module is used by the next module of the system. The existing systems are fully functional but not completely dependable. A user can perform all the actions of a library virtually in the existing management system. We observe that whenever a book is being stocked out of the library, the existing system only shows that which book is issued by the student or the teacher. But at the time of examinations it is very obvious that the particular subject books are very likely to be stocked out. At this time students issue more books of that subject, which results in stock out of the book in the library system, and there is no way that the other student who wants the book, can get the book in less period of time. As outcome of this scenario many students are left to study from the written notes all by themselves, without the help of any reference books. To eliminate this problem a stock analysis of the books can be generated which can show us the usage of particular books in a specific time period. This way a librarian can know which book are being extensively used in the time period and they can provide the book in the library after analyzing the stocks of the books. 1.2. Objective Here in the proposed work of ours we are implementing a library book-based analysis system using data clustering algorithms and data analytics techniques. We have the following goals for this project To provide a new mechanism for providing the stock analysis of the books available in the library database. To show which book has been issued mostly by the students as compared to the other books available for the same subject. To provide an analysis which will help the librarian to understand which books have the maximum uses which will help them to refill the stocks of those particulars books easily and to satisfy the need of a student or reader in time. 1.3 Scope This system will be effective in analysis of the books of the library as well as predicting the stock out of the particular books in the quarter of time. This will be done by utilizing the existing database of the library. The separation is done according to departmental basis i.e., user can generate the analysis after selecting the department, the semester of which the report or analysis is to be generated. This system is very much useful from the perspective of Librarian because in traditional library systems, there is a lack of module which can show the stocks of library books by predicting the usage of the particular books in particular time period. Organization of the rest of the sections are as follows, Section II describe the literature review Section III describes Problem Formulations Section IV Proposed System followed by References. 2. LITERATURE REVIEW Jian Wei Li et al. 1 proposed that with the constant development of modern library, the function of library has changed gradually. How to improve the utilization rate of library resources, how to serve readers better, and how to play more active roles, all have been becoming the concrete task of library in future. Clustering analysis is the process of grouping a set of physical abstract objects into classes of similar objects, and it has a very good application in library. The clustering analysis of readers behavior features in library automation system helps library improve services quality greatly, provide effective decision-making support for resource optimization, provide personalized information services for readers, and made library play more active roles in serving readers. 1 Tingting Zhu, Lili Zhang et al. 2 with the continuous development of librarianship, the functions of university library change. How to analyze the needs of university library users more effectively and rationally, thus provide corresponding service for the readers, has become a specific task which the future development of library will face. Data mining techniques can transform the collected data as questionnaires seeming to be uncorrelated and discrete into usable reference information provided to the library decision makers, which results in the effective dispose of the factors restraining users from using the library digital resource. 2 Ana Kovacevic et al. 3 With the increase in universal data volume, the technology of big data and its analytical processes are generally used to provide the description about massive datasets. Compared with other traditional datasets and its processes, big data includes semi structured and unstructured data that need more real time analysis. Big data also gets details about new prospects for determining new values, supports us to improve an in-depth understanding of the hidden values, and also incurs new challenges. 3 M.D. Anto Praveena et.al. 4 Since the field of Information Technology (IT) is improving a lot recently, this generates the data more easily. For instance, for every minute approximately 72 hours of video files are uploaded to YouTube by the people. This data growth challenges the field with the main problems of gathering and integrating huge volume of data from widely distributed data sources such as social media applications. 4 Raja Thangiah et al. 5 This study made an analysis of stock verification process in 12 academic libraries in Coimbatore district. This study reveals that, 66.67 percentage of colleges take stock verification through separate library committee from outside, 58.33 percentage of colleges take stock verification for the purpose checking the availability of books. 33.33 percentage of colleges lost 50 to 100 books at the time of stock verification,58.33 percentage of colleges take responsibility for the loss of books by the college management, 58.33 percentage of the colleges measure the loss through 3 percentages of total circulation of books in the library. 5 Fengjie Hao et al. 8 Digital library resources include text information, graphics, images, video, etc., mainly includes unstructured data, and partly are structured and semi-structured data. The data type is variety, and data generation, access and update are faster, meeting characteristics of the large data, velocity and variety. 8 Amir Michail et al. 6 Data mining is widely used in business to gain a competitive edge. An effective data mining application in the retail environment is shopping basket analysis. Progress in barcode technology has made it possible to store busker dura that contains items purchased on a per-transaction basis. By using data mining technology, one can find patterns in items that are bought in combination. 6 Wensheng Wang et al. 9 Web data mining decides which books are browsed most and which books should be booked largely. It gives a qualitative standard. But when to book, how many books should we book and what is the best safe stocks, these problems will be solved by optimization theory which can give a detailed booking quantity standard. 9 Maram Abdulrahman Almaghrabi et al. 7 We have defined the data mining problem and made an initial plan for recommending KOBSON services to new users by analyzing the process with domain experts and identifying their need to improve KOBSONs user-oriented services. Our objective was to help new users of the KOBSON DL (as well as the ones who have a problem in finding relevant resources) in finding appropriate information by recommending them the service that similar users have found the most useful for them. The recommendation can be generated starting from the users profile information and from the similarity of the users behavior (when interacting with the KOBSON DL) with that of other users. The data mining process is performed as illustrated in Figure 2. Users search history data (from the KOBSON proxy server log) and their profiles (from the KOBSON database) are collected and then analyzed. Appropriate data from the log are parsed using regular expressions. From the log file, we extract users who have downloaded at least one paper from the DL. The parsed log data and user profiles are then loaded into the previously created database tables (Oracle 10 g rel. 10.2, www.oracle.com), as shown in Figure 2. In the next step, data preparation, these data are transformed and normalized into a format suitable for clustering the users 7 Ahalya .G et al. 11 Every group is known as a cluster, which consists of objects that have affinity within the cluster and disparity with the objects in other groups. This paper is intended to examine and evaluate various data clustering algorithms. The two major categories of clustering approaches are partition and hierarchical clustering. The algorithms which are dealt here are k-means clustering algorithm, hierarchical clustering algorithm, density-based clustering algorithm, self-organizing map algorithm, and expectation maximization clustering algorithm. All the mentioned algorithms are explained and analyzed based on the factors like the size of the dataset, type of the data set, number of clusters created, quality, accuracy and performance. This paper also provides the information about the tools which are used to implement clustering approaches. 11 Prof. M. A. Deshmukh et al. 12 Cluster analysis itself is not one specific algorithm, but the general task to be solved. It can be achieved by various algorithms that differ significantly in their notion of what constitutes a cluster and how to efficiently find them. Popular notions of clusters include groups with small distances among the cluster members, dense areas of the data space, intervals or particular statistical distributions. 12 2.1 Background History Peng Hua Chen et al. 1 According to the statistical analysis, the departments of Chinese, politics and law, finance, foreign language and Mathematics are the most active departments whose students borrowed much more books than the other departments. But the numbers of students of the above departments are also much more than that of other departments. So, whats that could not reflect the objective demand of readers. From the results of cluster analysis based on the average loan rates of departments (book/person/year), we can see that The numbers of some departments are fewer, but the average loan rates of them are relatively high. 1 Lili Zhang et al. 2 Librarians not only should pay more attention to the specialties and academics research of all subjects, but also should not ignore the needs of some of small departments. And then, the library also can focus on recommending books to the active groups, have interactive communication with readers, and play more active roles to achieve the goal of efficient access to the reader needs and the reasonable books procurement based on the results of clustering. 2 Girija Chetty et al. 7 The dataset has subjects field corresponding to different disciplines, with each subject given a catalogue number. In addition, to text-based resources, different multimedia sources have also been included as resources. To the best of our knowledge, this is the first publicly available research testbed for investigating personalization and user centered modelling in the area of digital libraries and repositories. 7 Vladan Devedzic et al. 3 Classification algorithms may be tested with new records, based on the classification model built. The results obtained with the test data, can be used as an indicator, of how well the model will work with new data. 3 Wensheng Wang et al. 9 With the explosive growth of information sources available on the World Wide Web, it has become increasingly necessary for users to utilize automated tools in find the desired information resources, and to track and analyze their usage patterns. These factors give rise to the necessity of creating server side and client-side intelligent systems that can effectively mine for knowledge. Web mining can be broadly defined as the discovery and analysis of useful information from the World Wide Web. This describes the automatic search of information resources available online. 9 Prof. R. A. Gulhane et al. 15 The appropriate clustering algorithm and parameter settings (including values such as the distance function to use, a density threshold or the number of expected clusters) depend on the individual data set and intended use of the results. Cluster analysis as such is not an automatic task, but an iterative process of knowledge discovery or interactive multi-objective optimization that involves trial and failure. 15 2.2 Existing System Shaochun Xu et al. 10 Libraries have collected a large amount of data, such as books, research articles and reports, both in physical and electronic formats. The library collection was originally for researchers or public users to find necessary information they need. However, this data becomes so large and the format is so various which might affect the efficient use. Although a lot of library data has been digitalized, most of them have not been used for data mining or big data technology. 10 Wensheng Wang et al. 9 On the other hand, although some work has been done in the past on how to maintain those library collections in order to efficiently and effectively use. there is no much research on using meta data to organize digital assets so that the big data and cloud computing technology could be used. 9 Ksm Swaminathan et al. 5 In the context of libraries, verification of stock is different from the verification of stock in stores in Private or Government business organizations, the purpose of the job and the difference in the meaning of Store and Stock are concerned. The collection in the libraries contains various types of information sources. The library is a public institution and collections subjected to accounting and checking, verification and reporting. 5 Raja Thangiah et al. 5 Most of the libraries have open access and there are chances to damage, misplacement and loss of books. But this system should be thrown away by the slogan by the father of library science Dr. S.R.Ranganathan, Books are for use. The books should be placed in shelves in proper way to identify by the readers and library staff. Hence, the stock verification becomes easy and important for library activity. In an open access library reader can handle the books freely in the stock room. 5 Amir Michail et al. 6 Large amount of data is gathered in different databases because of advanced methods of data collection. The demand for grouping the important data and extract the useful information from data is increased. Clustering is the distribution of data into groups of identical objects which has affinity within the cluster and disparity with the objects in the other groups. Patterns within the same cluster are closely related than to the data in the adjacent clusters. Here, there is need to know the difference between the unsupervised classification and supervised classification that is between clustering and discriminate analysis 6 Maram Abdulrahman Almaghrabi et al. 7 The model structured by Decision Tree C4.5 Algorithm in the data mining, the author gains some potential links among the factors affecting the using of library digital resource. This kind of mining information through decision tree algorithm can enable the library decision makers to know about the users various needs of using the librarys information resource, so it provides scientific reference indexes for the librarys future reform and development, and provides sufficient and credible data for the follow-up improvement of service. In other words, it furthest increases the use ratio of library resources, playing a scientific analysis and prediction role for the development of librarianship. Certainly, because the sample size and items in this model are not enough, analyzing degree and knowledge contents gained are not satisfying, and the credibility is not high enough, this model can only be named as a simple and shallow classification analysis model.. 7 Lichao Chen et al. 10 The hardware and software for storing and analyzing big data is cheaper and available for business and government now which makes the big data technique interesting to a lot of users including library. The important part is that the user could make prediction based on big data analysis. Work about big data in library could also be found because library data need to be transformed into information or knowledge which then be used by users. Bell tried to explore the issues and possibility of big data in library. Parry studied how colleges are using big data to help students chose classes, retain them, and provided necessary advising. 10 Amir Michail et al. 6 ways library classes are reused in practice, which we call reuse patterns. This is done by data mining existing applications that use the library. For example, we may find that most application classes that inherit from a library class Widget tend to override its member function paint(). As another example, we may notice that most application classes that instantiate a library class Painter and that call its member function begin() also call its member function end(). 6 2.3 Limitations Jian Wei Li et al. 1 In traditional library, the decisions are subjective, one sided or blind most of the time because the decision-making rely on the experience, then they could not adapt to the development times. The unified focus decision-making information could be extracted from a variety of internal and external information involved in the Library Information. 1 Lili Zhang et al. 2 Due to lack of technical support means, the former service promotion strategies of the library were mostly established based on experience. As a result, the propagandas were not niche targeting at all, which greatly weakened the promotion effect. Through analyzing the user survey questionnaires with the help of the decision model, we reach a conclusion that whether the readers know about the librarys digital resources is the most direct factor that influences the use ratio of digital resources. 2 Viktor Pocajt et al. 3 Classification algorithms may be tested with new records, based on the classification model built. The results obtained with the test data, can be used as an indicator, of how well the model will work with new data. We used 60 per cent of our data for building the model and the remaining 40 per cent for testing it. 3 Dr. B. Bharathi et al. 4 Many datasets have definite levels of heterogeneity in structure, semantics, type, organization, granularity and accessibility. Data representation aims to make data more important for data analytics and user analysis. Any improper data representation may reduce the value of the data originality and even disturbs effective data analysis process Hence if the data is represented effectively, then analysis process will be done easier. 4 Girija Chetty et al. 7 The dataset has subjects field corresponding to different disciplines, with each subject given a catalogue number. In addition, to text-based resources, different multimedia sources have also been included as resources. The methodology for the design of the data store is as shown in the Figure 1 schematic. To the best of our knowledge, this is the first publicly available research testbed for investigating personalization and user centered modelling in the area of digital libraries and repositories. 7 Hari Mohan Pandey et al. 11 Characterizing data into a smaller number of clusters will definitely lead to a loss in some details but data will be interpreted. It represents data objects by fewer numbers of clusters and thus, it models, data by using its own clusters. Cluster analysis is the arrangement of a set of patterns (generally shown as a vector of measurements multidimensional space) into clusters based on similarity. The explosion of availability of information on the Internet requires that libraries evolve into value-added information providers, rather than mere curators of collections. Modern libraries need to stay relevant to a diverse, technologically savvy patron base and to facilitate and add value to the research community, while facing significant resource constraints. To face these challenges, libraries need to embrace digital technologies and library management systems (LMS) in order to work smart and achieve more with less. While LMS systems have been around for decades, libraries can explore the new frontier by embracing open source solutions, like open source software (OSS) library systems, which are free to acquire. They need to collaborate with computer experts and become technologically savvy to harness the full power of OSS solutions to meet the specific needs of the library and patron base. This article provides an overview of the availability, benefits, and drawbacks of various LMS systems and OSS variants, drawing from experiences in the present Indian context. 11 3. PROBLEM FORMULATION In the traditional library systems numerous fields are being considered from the perspective of the administrator and the user of the system. Various fields define the system. Data mining techniques are used to recommend digital library services based on the users profile and search history. First, similar users were clustered together, based on their profiles and search behavior. Then predictive classification for recommending appropriate services to them was used. It has been shown that users in the same cluster have a high probability of accepting similar services or their patterns. Data mining is widely used in a system now a days because as the databases are rapidly increasing day by day, there is a much need of the data analytics tools because if the smallest amount of data loses its integrity in the database, there may be chance of occurrence of false results and decisions in the algorithmic instructions. Following system comprises both the data mining techniques and data analysis algorithms which is crucial for the system architecture. Our system implements a method to analyze and generate a stock verification which can represent usage of particular book stock in a selected time period. 3.1 Aim Our aim is to develop a system i.e., Library Book Based Analysis System, which is based on the Data Mining techniques and as well as the Data analytics technologies. We are using the existing Library Database of Our College. This way we are keeping our project relevant to the existing Database. The main aim of our project is to analyze and represent the stock of the books of our college library, so that the librarian can know which books are more likely to get stock out in that time so that they can arrange more books of that particular genre in our library. So that the students can get the books in proper time as they required. This system will also show that which books are not being used frequently by the students, so that library can update new books in place of those old least used books. 3.2 Problem Definition In library, readers have variable behavior characteristics in borrowing books. They can issue whichever book they want from the library. The existing system stores entry if that particular book being issued by the reader, and the stock is decremented by the number of particular books being issued. This is done by the scanning the bar code of the library book which responds to the particular book data available in the database. The observation is that, whenever a book is issued, the existing system cannot predict that which book is going to be used the most in upcoming time. It can only show that which book is stock out at the current moment. This results in students not getting the required book whenever they need it. This scenario is much likely to be seen in every library of any college. Our system will be designed in such a way that it can predict and analyze the book being used extensively in a selected timeline by using the existing library database to keep the system relevant as much as possible to the current scenario. 4. PROPOSED SYSTEM In traditional library, the decisions are subjective, and predictive if proper approach is used or applied onto the existing system. It is one sided or blind most of the time because the decision-making relies on the experience, then they could not adapt to the development times. In the proposed system we will use the datasets obtained from the library, which is utilized for the analytical purpose using various association mining, Data mining and Data Analysis techniques. The proposed system will show usage of books, books status, and stock variation. The system will be divided into various modules which will function out various calculations onto the Database i.e., it can be association mining, clustering, or analysis of the data. The Filtration of the departments will be done for the easy approach for the system. After selecting the department and semester user will get a list of books where he can easily find out that which book is having the maximum usage amongst the other books of that particular subject. The proposed shows a stock analysis which also helps library to find out the theft or misplacing of the library books. The use of numerous techniques in the system makes this system depended as itll be more efficient to use and analyze the Datasets. Each variation in stock is shown in a graphical manner, so user can easily see the analysis reports easily. This report is generated using the J-Free Integration Module of java.. 4.1 Methodology As the time progresses, the functions of university library change. How to analyze the needs of university library users more effectively and rationally, thus provide corresponding service for the readers, has become a specific task which the future development of library will face. This research is done by the survey method and the questionnaire is used as a tool. The questionnaire method was followed to get data for the study. The stock verification becomes easy and important for library activity. In an open access library reader can handle the books freely in the stock room. The stock verification has some major advantages in the stock verification including disclose the position of the loss of documents, so that the replacement may be made in case of important document loss which provides adequate account of the percentage of inevitable loss and provides opportunities to weed out long un-used, very old editions of books from the library. The methodology can be understood in following ways- After defining the project objectives and requirements, formulate data mining problem definition and prepare initial strategy for achieving these goals. After collecting the data, analyze it to familiarize with it and discover initial insight also evaluate the quality of data. Prepare the final data set from the initial raw data that will be used in the process. Select cases and variables that are appropriate for analyses and perform the necessary data transformations. Apply appropriate modeling techniques and calibrate the model to optimize the results. The results show patterns (i.e. the model) discovered for the data analyzed. If necessary, loop back to the preparation phase to bring the form of the data in line with the specific requirements for particular data mining techniques. 4.2 Algorithm Following Algorithms and techniques are used in the proposed system- Apriori Algorithm Apriori is an algorithm for frequent item set mining and association rule learning over transactional databases. It proceeds by identifying the frequent individual items in the database and extending them to larger and larger item sets as long as those item sets appear sufficiently often in the database. The frequent item sets determined by Apriori can be used to determine association rules which highlight general trends in the database this has applications in domains such as market basket analysis. Apriori uses a bottom up approach, where frequent subsets are extended one item at a time (a step known as candidate generation), and groups of candidates are tested against the data. The algorithm terminates when no further successful extensions are found. Apriori uses breadth-first search and a Hash tree structure to count candidate item sets efficiently. It generates candidate item sets of length k from item sets of length k-1. Then it prunes the candidates which have an infrequent sub pattern. According to the downward closure lemma, the candidate set contains all frequent k-length item sets. After that, it scans the transaction database to determine frequent item sets among the candidates. Method Let k1 Generate frequent item sets of length 1 Repeat until no new frequent item sets are identified 1. Generate length (k1) candidate item sets from length k frequent item sets 2. Prune candidate item sets containing subsets of length k that are infrequent 3. Count the support of each candidate by scanning the DB 4. Eliminate candidates that are infrequent, leaving only those that are frequent Furthermore, the Apriori algorithm can be understood by given pseudo-code Join Step Ck is generated by joining Lk-1 with itself. Prune Step Any (k-1)-itemset that is not frequent cannot be a subset of a frequent k-itemset Pseudo-code Ck Candidate itemset of size k K-means is a famous non-supervised clustering algorithm used to organize the data. It is a partitioning algorithm wherein the resultant clusters are independent and bound. There are broadly two main stages of algorithm implementation. Firstly, the data is divided into k number of clusters with assumed k value in advance. Among the given set of data take k number of points and assume it as a centroid for that respective cluster. Secondly, calculate the distance between point and centroid and assign the point to the cluster, which has the least distance which brings the point closest centroid. This method reduces the number of iterations and change of locations of points in clusters. K-means (MacQueen, 1967) is one of the classical unsupervised learning algorithms that solve the well-known clustering problem. The procedure follows a simple and easy way to classify a given data set through a certain number of clusters (assume k clusters) fixed a priori. The main idea is to define k centroids, one for each cluster. These centroids should be placed in a cunning way because of different location causes different result. 1. Place K points into the space represented by the objects that are being clustered. These points represent initial group centroids. 2. Assign each object to the group that has the closest centroid. 3. When all objects have been assigned, recalculate the positions of the K centroids. 4. Repeat Steps 2 and 3 until the centroids no longer move. This produces a separation of the objects into groups from which the metric to be minimized can be calculated. Given Ki ti1, ti2 ,, ti(m), the centroids function Let us understand this equation by example, suppose the tuple is needed to be clustered is 2, 4, 10, 12, 3, 20, 30, 11, 25 given k2 and then select m12 and m24 as the initial group centroids. The distance between point and centroid is measured by Euclidean Distance The results of cluster analysis of the tuples by implementing k-means algorithm as shown in Table Table 1.1 Obtained Value of Clustering Example Number Of Iterations M1 M2 K1 K2 1 3 18 2,3,4,10 12,20,30,11,25 2 4.75 19.6 2,3,4,10,11,12 20,30,25 3 7 25 2,3,4,10,11,12 20,30,25 C) Association Mining Rule Learning Association rule learning is a rule-based machine learning method for discovering interesting relations between variables in large databases. It is intended to identify strong rules discovered in databases using some measures of interestingness. This rule-based approach also generates new rules as it analyzes more data. The ultimate goal, assuming a large enough dataset, is to help a machine mimic the human brains feature extraction and abstract association capabilities from new uncategorized data. Association Rule 4.3 Flow Chart Figure 1.1 Flow Chart for Analysis System 4.4 Data Flow Diagram Level 0 Data Graphs Figure 2.1 Level 0 Data Flow Diagram Level 1 Figure 2.1 Level 1 Data Flow Diagram 4.5 UML Activity Diagram Figure 4.1 Activity Diagram for Analysis System Used Case Diagram extends User Fig 5.1 Used Case Diagram for Analysis System REFERENCES 1 Jian Wei Li, Peng Hua Chen, The application of Cluster analysis in Library system. Institute of Electrical and Electronics Engineers (IEEE) 978-1-4244-3531 February 2008. 2 Tingting Zhu, Lili Zhang, Application of Data Mining in the Analysis of Needs of University Library Users, The 6th International Conference on Computer Science Education (ICCSE 2011). Superstar Virgo, Singapore. August 3-5, 2011. 3 Ana Kovacevic, Vladan Devedzic, Viktor Pocajt, Using data mining to improve digital library services, ResearchGate, 05 June 2014. 4 M.D. Anto Praveena, Dr. B. Bharathi, A Survey Paper on Big Data Analytics, International Conference on Information, Communication Embedded Systems 2017. 5 Raja Thangiah, Ksm Swaminathan, Stock Verification of Books in Academic Libraries A Special Reference to Selected College Libraries in Coimbatore District A Study, ResearchGate, 15 January 2018. 6 Amir Michail, Data Mining Library Reuse Patterns using Generalized Association Rules, Institute of Electrical and Electronics Engineers (IEEE), March 2000. 8 Fengjie Hao, Fei Liu, Research of Hadoop-based digital library data service system, 9th International Conference on Intelligent Human-Machine Systems and Cybernetics, 2017. 9 Wensheng Wang, A Library Booking Policy based on Stocking Theory, IITA International Conference on Services Science, Management and Engineering, 2009. 10 Chunning Wang, Shaochun Xu, Lichao Chen, Xuhui Chen, Exposing Library Data with Big Data Technology A Review, Okayama, Japan, 2016 IEEE ICIS 2016, June 26-29, 2016. 11 Ahalya .G, Hari Mohan Pandey, Data Clustering Approaches Survey and Analysis. 2015 1st International Conference on Futuristic trend in Computational Analysis and Knowledge Management (ABLAZE-2015), 2015. 12 Prof. M. A. Deshmukh, Prof. R. A. Gulhane, Importance of Clustering in Data Mining, International Journal of Scientific Engineering Research, Volume 7, Issue 2, February-2016 PAGE PAGE MERGEFORMAT 2 PAGE MERGEFORMAT 2 Library Analysis User User Library Analysis User Data Preprocess Data Analysis Graph Manager Database Load database to system Preprocess dataset from the acquired database Perform k-means clusters for semester and department wefijfjwfdededddddepardddddepartmentdepartment Perform analysis for individual books and groups (year,department,semester) Perform graph-based analysis ( x co-ordinate stock of book available and y co-ordinate time in which book is issued ) Load Data Analysis Perform Graphing Perform Apriori Perform Clustering Process Dataset
xxmpmeta xmlnsxadobensmeta/rdfRDF xmlnsrdfhttp//www.w3.org/1999/02/22-rdf-syntax-nsrdfDescription rdfaboutuuidfaf5bdd5-ba3d-11da-ad31-d33d75182f1b xmlnsxmphttp//ns.adobe.com/xap/1.0/xmpCreateDate2018-10-12T164900.176/xmpCreateDate/rdfDescription/rdfRDF/xxmpmeta kWe4Yi2qACdrWO SE
xxmpmeta xmlnsxadobensmeta/rdfRDF xmlnsrdfhttp//www.w3.org/1999/02/22-rdf-syntax-nsrdfDescription rdfaboutuuidfaf5bdd5-ba3d-11da-ad31-d33d75182f1b xmlnsxmphttp//ns.adobe.com/xap/1.0/xmpCreateDate2018-10-12T164917.916/xmpCreateDate/rdfDescription/rdfRDF/xxmpmeta jy2IZS8m5a-d4wX5DRy68pr 4Pz s7 w5ZMK kB3rwWg 0w3CDO mmFcrRIJs BtXwcWVmL73USJ-tLZDurZzQR0_vmvjZw3AI1iGVtLDQ4SFNYI1gts. fT15lzp1Ze_
xxmpmeta xmlnsxadobensmeta/rdfRDF xmlnsrdfhttp//www.w3.org/1999/02/22-rdf-syntax-nsrdfDescription rdfaboutuuidfaf5bdd5-ba3d-11da-ad31-d33d75182f1b xmlnsxmphttp//ns.adobe.com/xap/1.0/xmpCreateDate2018-10-13T171307.320/xmpCreateDate/rdfDescription/rdfRDF/xxmpmeta 0x9fucCEm-/E5)8u_ PeEsouj4xY V u7J 16hmY4wdNBj6D HgW(9(EhAUd)XAl f7to31ocPLl9
[email protected]@D0I//1jpG)Qu FVf,4Znp8dq6upwoqnX2j883Avm5 BuVoy04.84 DqlPwD5Wo X 4n3)RPo-UXE5I7
qOC.9iYBTyqcIMkRF4ZK7Z
J_9VuMuE,YU6pXXuO
kRI9TADNP 7EZNEnKnIVI/ZKAd 1.PqoBHCB_ZLn-OCQh4YSSIf1XnyW 4 /i [email protected],,-BgzpPBojZkKFZMc6
@0u RTIvYMN/u0X T3V-sVmIky T -Gq3b,k [email protected] 2yum7Xk/gPRO_ajYMCYAkx,OzKwtZiy_cS3a4TIJKA3OdF()YJu_XW9F2sMCs,7Za2A HcvCjJ_x [email protected] eeOVnIax.y81wd8(7OZ5MkAFN2x
[email protected]//pBMkW-MY2k
,ZL(pO-W mhEoRThyLm59OckZYG02)KaQ2@(-fiAnnfgqVMoAz74OIdxI7 sXuI5MRR 4FFI ig
r1YbkkDG.JmAc/_tVD u7J kzV5gc-vsp6b7GWxG_ r _iOiij)H6psciq6I Ox8o.yJrHdAe kzMIIcfB
ziu8M _xLVzEwZy,@ Yktrrx4 7jY6bf.IMyPLgPexr
)zz/1vGkmxmb.W [email protected]/[email protected],n0Q_-muk,3Fg9WjGS-hnh0R3K/ Gp,CYGqRj76zM.bJ upgideXjERVLNXZ M9NMH(C
(
(
()jk-,y 0Tmkze/
m,RhorG Y, B8L 1(IzZYrH9pd4n(KgVB,lDAeX)Ly5otebW3gpj/gQjZTae9i5j5fE514g7vnO( ,[email protected] /[email protected] 6Q 790( g,JNSQp7otykEiMEJfsJBdfQjeJ1UAL(W6FWE,XEuU1LNyEt7iDUnY,.YeB3i5 M.e0LoFZ nw3Yyghg X/[email protected] gzSj8YJv ,0t9De)FC 2DfZ @/jpbB1(fa,lw, twD5Y,.Xf kzx.lI.r vJ mpb9NG
iPtKI
xXghmP.W7 ozw5NL.FWEz IE 9eV)qSQ(g87kEviMMpHsJBdaQrE1UNLzW.FWE,[email protected](k_.SHmzVh@ 7/PzeW ev_zd3GYFXkn-kE879eNkJezdReGCj)n(- 7/wY.Z)tP8005w9vzAEuss4d6 fjf4Jw moV1prSl7 du)MoAOR/JWfee RS8LNCZSecWPbzUoDF/8wgbWq0JndmjIBg
v.XU(s1K-Kr , mCet2QtZ3s8 eY.n58Y
v z v2Ctq L)b
5(ZuE.R35SE_WmbB1(ofal,BCthtt.BUoiMDkp UM 0DIcH(2P Of8IMCe F)bsq8zR/7
bVML gVJi ),6tU-s7I48D5S(@(p.nxf6a
2 [email protected] @cLtEjsGHmX(yOgPc7SmOy3e
G qmWBDyQXX7/FI 41naXnk82mH4I2eOw xvp9hxs pgFC0K,Ok@ rci YB EelMw Jo_dmEo(5RCtK6Cn Qt)s jQ aYf,B6)S Qr)ojXHqOY)gvXSpeL8J9pBfdy/sT
32WiiB.E mGG AIoNNd1tnB CXKABYbqDHvsmZfkU3acBrUj,yTr6 C8Zua
6pkdAcee wrY0XH-CXefbzPioDF/Z8W7U12_dD) L6Jp@
Hcs63Re97mVf Hg.mtE)
3C