Abstract:

Modern trends in computer systems are associated with the development of high-performance computing and working with databases with an increasing volume. The vast amount of collected data significantly exceeds human’s ability to effectively benefit from them, without the help of specialized tools for data analysis. In this situation, the urgent question is how to avoid the well known situation “rich in data, poor in information”. Processing and storage of large amounts of data requires a new perspective and joint implementation of a number of established technologies. A new concept – ‘Data Science’ – has appeared, which is associated with retrieving previously unknown knowledge directly from the data through processes of detection and analysis of hypotheses. In response, as well as to ensure well-trained software professionals, universities have introduced a new discipline, or a new module, in the course “Databases”, associated with the application of techniques for data analysis.

The article presents the experience gained by applying Data Mining techniques to extract new patterns of data in the courses “Databases and knowledge” in some Bulgarian universities. The main emphasis in this training is the introduction of an adequate algorithm for extracting knowledge through a specific task. Solving the task shows how to extract new knowledge, describe the relationship between data properties, data models and others. It is pointed out what types of dependencies can be found in the process of extracting data patterns. To this end, we review different types of tasks – Data Description and Summarization, Cluster Analysis, Associative Analysis, Outlier Detection, Classification and Forecasting, etc.

Some problems identified during the training are analyzed. In addition, opportunities to improve learning results of the course “Databases” are discussed.

Keywords:

Data Minimg, Database, Big Data, Data Science, Education Space.