Hierarchical Visual Event Pattern Mining and its Applications

Questions: Evaluates interactive database systems, Explains the process of data mining, Explains online analytical process (OLAP), Assesses how visual data are extracted from image databases, Explains decisions are made using this data, Explores the professions that will most likely benefit from the information and knowledge created via visual data mining.? Answers: Interactive Database Systems: Day by day the technology is giving more new ways of finding refined solutions for critical types of works. Relational database systems are getting more popularity for this reason. In the interactive database system, the database is connected to the user via a device with graphical interface. The main goal of the system is to provide an interface of the database system which can be easily handled by the users (Feng et al., 2010). The typical way of handling the database management system is to use the SQL queries. The structured formulas of the SQL queries are not easy to handle by the users who are not professional in the field of database. Therefore, the interactive database system is very much needed for the efficient handling of the database system for the users in different areas. This system allows the users to handle the database without knowing the internal complex structure if the database management system. Data Mining Process: The concept Data Mining refers to the process, through which data can be analyzed from different perspectives and can be summarized into usable information. Data mining software is a kind of tools or set of tools which are useful for the purpose of analyzing data. The process gives the ways of analyzing data from different angles, categorizing the data, and summarizing the relationships among the data. In technical speaking, data mining is the process of finding the correlations among the data of a large database (Cui et al., 2010). The concept of the data mining process is comparatively new in the field of computing technology but has the power of providing great potential to help the companies for focusing on the important information of their data warehouses. Data mining system can be developed by using the existing software and hardware for the enhancement of the existing facilities in the field of database management. The tools of this system has the ability of providing answers of the business related questions which are too time consuming for solving in traditional technique. Actually the name data mining came from the concept of finding important data or information from a large volume of database. A large database consist a huge numbers of rows and columns (Ward, Peng Wang, 2004). A data mining system with high performance gives the chance of exploring the database in full depth, without selecting a subset of the data stored in the database previously. There are six techniques of the data mining system available that are widely used. They are: artificial neural networking system, use of the decision trees, genetic algorithms, nearest neighbour method, the method of rule induction, and the method of data visualization. On-Line Analytical Process (OLAP): The On-Line Analytical Process (OLAP) is getting more popularity in the field of the data warehouse. It is a technology that is currently using for managing the large database of the business organizations. This system provides an online facility for the database management and queries related to the business-intelligence. In the OLAP system, the necessary data are stored in data warehouses (Georgieva-Trifonova, 2011). This is a hierarchical database system where the data are stored in some cubes rather than the tables of the traditional database management system. In this system, the databases are divided into some cubes. Each of the cubes is designed and managed by individual cube administrators. Actually this system is developed for providing high rapid access to the database with a multi dimensional architecture. Basically there are two types of data in the database of the OLAP system. They are measures and dimensions (Hsu Li, 2011). Measures are numeric type of data which are b asically the quantities and averages of the information related to the business decision. Dimensions are used for categorizing and organizing the measures. The OLAP system has the following sections: Cube: This is a data structure for maintaining the hierarchical architecture and level of the OLAP system. The measures are stored in the cubes with several dimensions like time, product line, geography etc (Patel Patel, 2011). Measure: Measures are the pre-processed, analyzed and aggregated values that are stored in the Cubes. The common examples of the measures are profits, costs, sales revenue etc. Member: The occurrences of the data are represented as the members. The members can be either unique or non-unique. As an example, under the dimension of time 2008 and 2009 are unique members of the year level but January and February are non unique member of the month level. Calculated Member: The value of which member of a dimension is calculated at runtime through an expression is a calculated member. An example of the calculated member is profit as the value of profit is determined by subtracting the value of cost from the value of sales (Zhang Shen, 2010). Dimension: These are the set of levels which are hierarchically organized in the cubes for understanding the database and analyzing the data. Hierarchy: This is a logical tree structure for organizing the members of the dimensions. Level: In the hierarchical structure, data can be organized into higher and lower level. As an example year, month and day are the different types of levels in the time hierarchy. Extracting the visual data from image databases: The important features of the image data are texture, shape and colour. The extraction of the visual data from the image database deals with the extractions of these three features. Texture extraction: Texture of the visual data is an important factor for the human vision. Different types of images have the different types of textures. Statistical and structural are the two major approaches of extracting the textures of the visual data. Texture based queries are used for the extraction of textures. In this case the query provides images that are having similar types of textures (Mahbubul Majumder, 2013). Shape extraction: Shape of an image file is extremely important in the field of the visual data. The traditional approaches that are used for the shape extraction are edge based, feature based and region based shape extraction technique (Necir, 2010). Colour extraction: Extracting the colour features of an image according to the human vision is not an easy task. Major approaches used for the colour extraction are single colour, colour pair and the histogram of the colour images (Yu, Yurovsky Xu, 2011). Decisions made using this data: Decision making is the most important task of any business. An efficient data visualization tool is very important for the decision making. It gives the opportunity of taking the business related decisions quickly by examining the large amount of data. The practice of using the visual data is not a new technique. The visual information is used for the decision making by the scientists, students, business analysts. Visual data that are measurable give the scope of easily examine and point out the outliers to the decision makers. The use of the visual data for making the decisions is very popular but the available strategies are very poor. The key factor behind getting success in decision making by using the visual data is to use the appropriate type of the data visualization. A good strategy of data visualization gives the opportunity of taking the decisions more quickly and confidently. Professions that are getting benefit from the information and knowledge created via visual data mining: Mostly the business organizations are getting benefits from the visual data mining. The database of the business organizations contains a large amount of data. Using the concept of the data visualization is very helpful for organizing the database and taking decisions using the data stored in the database. Information regarding the marketing and the business management can be efficiently handled through the visual data mining. The scientific research works are getting a huge benefit from this concept. These works also needed large amounts of important data stored in the database. Storing the visual data regarding the important information is giving the ability of doing further retrieval and understanding of those data easier and more efficient. References: Cui, P., Liu, Z., Sun, L., Yang, S. (2010). Hierarchical visual event pattern mining and its applications. Data Mining And Knowledge Discovery, 22(3), 467-492. doi:10.1007/s10618-010-0195-5 Feng, S., Zhao, Z., Zeng, Q., Fan, J., Zhang, X. (2010). Personalized Knowledge Acquisition through Interactive Data Analysis in E-learning System. JCP, 5(5). doi:10.4304/jcp.5.5.709-716 Georgieva-Trifonova, T. (2011). Warehousing and OLAP Analysis of Bibliographic Data*. IIM, 03(05), 190-197. doi:10.4236/iim.2011.35023 Hsu, K., Li, M. (2011). Techniques for finding similarity knowledge in OLAP reports. Expert Systems With Applications, 38(4), 3743-3756. doi:10.1016/j.eswa.2010.09.033 Mahbubul Majumder, T. (2013). Visual Mining Methods for RNA-Seq Data: Data Structure, Dispersion Estimation and Significance Testing. Journal Of Data Mining In Genomics Proteomics, 04(04). doi:10.4172/2153-0602.1000139 Necir, H. (2010). A data mining approach for efficient selection bitmap join index. International Journal Of Data Mining, Modelling And Management, 2(3), 238. doi:10.1504/ijdmmm.2010.033535 Patel, A., Patel, D. (2011). Multidimensional model and OLAP operations. IJAR, 3(3), 57-58. doi:10.15373/2249555x/mar2013/20 Ward, M., Peng, W., Wang, X. (2004). Hierarchical visual data mining for large-scale data. Computational Statistics, 19(1), 147-158. doi:10.1007/bf02915281 Yu, C., Yurovsky, D., Xu, T. (2011). Visual Data Mining: An Exploratory Approach to Analyzing Temporal Patterns of Eye Movements. Infancy, 17(1), 33-60. doi:10.1111/j.1532-7078.2011.00095.x Zhang, Y., Shen, X. (2010). Model selection procedure for high-dimensional data. Statistical Analysis And Data Mining, 3(5), 350-358. doi:10.1002/sam.10088