1.1 Summary of Data Mining
Data mining is a technology that operates in large databases or data warehouses. Its purpose is to analyze useful information hidden in large amounts of data. With the continuous development of data acquisition and storage technology, a large number of data is stored in databases in various fields, and valuable information is obtained from these massive data, which ultimately forms the theme of data mining. Therefore, the ultimate goal of data mining is to help analysts discover some relationships between data after data mining, and find hidden information easily overlooked in large amounts of data, which is usually conducive to industrial decision-making and behavior prediction (Mengmeng, Weiwei, & Honglin, 2019).
Data mining uses probability theory, artificial intelligence, machine learning and other technologies to analyze data in file system and database system, gains a certain degree of trust, and supports decision-making process from different perspectives.
In the era of big data, data mining needs to be built on the premise of data collection and organization. For large data, the role of data mining is very important. Generally speaking, data mining, also known as knowledge discovery, is defined as an engineering and system process to explore hidden, previously unknown and potentially useful information and patterns. The attributes of the process are related to the following aspects (Jiwu, 2019).
- 1.
Applicability: Data mining realizes the organic integration of algorithm and practice. The emergence of data mining conforms to the needs of the application and development of production process. The latter is also responsible for providing data, and completes the acquisition of useful information by means of data mining, and then puts it into practice again to provide help for decision-making. In this sense, data mining comes from practice and promotes practice.
- 2.
Engineering: Data mining has the nature of engineering, covering specific steps. According to the application characteristics, data mining is different from algorithmic analysis and application. It includes data preparation and management, data preprocessing and transformation, mining algorithm development and application, result display and verification, and knowledge accumulation and use.
- 3.
Collectivity: Data mining contains a variety of functions. Generally, data mining can complete data exploration and analysis, association rules mining, time series pattern mining, classification and prediction, clustering analysis, anomaly detection, data visualization and link analysis. In the process of practice, more than one function is used, and the corresponding functions involve specific principles and techniques, as well as complex algorithmic tools.
- 4.
Crossing: From the professional nature, data mining has cross-cutting characteristics, involving self-statistical analysis, pattern recognition, machine learning, artificial intelligence, information retrieval, database and so on. It is a combination of many disciplines. In addition, such as stochastic algorithm, information theory, distributed computing and other research, also provides a lot of enlightenment for the study of data mining. Compared with the above areas, data mining has its unique features, namely, the above-mentioned applicability, engineering and collectivity, especially in the application.