2.1 Association Rule Mining
Data mining is a hot spot in the research field of data mining technology and artificial intelligence technology, the so-called data mining is to reveal the previously unknown and potentially valuable hidden information from the data. Data mining is the most cutting-edge discipline that spans a variety of technologies, from simple query of data to the application of data analysis and statistics, from which useful information can be extracted to help planners change market strategies, reduce losses and make wise judgments.
Data mining tasks can generally be divided into: description and prediction[11].Descriptive data mining: express data based on a meaningful and concise overview of the nature of existence. Predictive data mining: analyze the model or a set of data obtained by providing settings to apply special data, and the model is new data used to predict the nature of the future.
Association rule mining is one of the most active data mining methods[12]. Association rule mining is the process of discovering frequently associated data items from a database. The role of association rule mining is to discover the relationship between items from the data set so as to guide decision-making. The mining process of association rules mainly includes two stages: the first stage must first be from the high-frequency item sets in the data, and the second stage must generate association rules from a group of high-frequency items.The first stage of association rule mining must be set from the original data[13], discover all frequent itemsets. The second stage of association rule mining is to generate association rules. The threshold conditions under the minimum confidence are set to use the high-frequency k-item generation rules in the previous step. If the reliability of the rules meets the minimum confidence expression, they are called association rules. Then set the frequency set found in the first step to generate ideal rules, and the result can only contain the set of all rule items, with only one right part of each rule. Once a defined rule is generated, users with a minimum confidence greater than the specified unique rule will be excluded.
2.2 Related algorithms of association rule mining
Apriori algorithm is an algorithm for mining frequent itemsets based on Boolean association rules. Frequent item sets refer to the existence of some identical elements in a series of sets, and the elements that appear frequently in the sets at the same time form a subset, which satisfies a certain threshold condition. The core of Apriori algorithm is the idea of realizing frequency sets based on two-level recursive algorithm. Association rules are the close association rules between item sets given by frequent item sets. The association rules in classification are single and one-dimensional Boolean association rules. When the support exceeds the user-defined threshold, the item set is called frequent item set, or frequency set for short.
The basic idea of this algorithm is to find out all the frequent sets, which are at least the same as the predefined minimum support count. Strong association rules are generated in frequent itemsets, and these rules must meet the minimum support and minimum confidence. Then set the frequency set found in the first step to generate ideal rules, and the result can only contain the set of all rule items, with only one right part of each rule. Once a defined rule is generated, users with a minimum confidence greater than the specified unique rule will be excluded.
The disadvantage of Apriori algorithm is that there are too many combinations of candidate item sets, and unnecessary parameter combinations must be removed. In the support calculation, the computer will scan all the data. Assuming that it is a huge database, the Apriori algorithm will increase the load of the computer system and increase the load in equal proportion. The flowchart of Apriori algorithm for association rules is shown in Fig. 1.
2.3 Microsoft SQL Server 2016 Overview
Microsoft SQL Server 2016 is an embedded database programming software, which simplifies the management and maintenance of the database. SQL can fully support web-based XML and Internet standard database programming, and send requests to access Web data and database applications through HTTP.SQL scalability and security can meet the requirements of large enterprises (such as Alibaba's global e-commerce enterprises and Foxconn's large enterprises).
SQL Server is a relational database management system, which is used to store and retrieve data. It can perform queries, store and retrieve data, update data in the database, and write or retrieve data from the database through SQL statements. The program development interface of Microsoft SQL server is friendly and the graphical interface is the same as that of Microsoft Windows, which is convenient for beginners.
In addition, as an application of the background database [14], Microsoft SQL Server 2016 must establish a connection with the foreground, and issue commands in the foreground to call the data in the background database or edit the data in the database. Microsoft SQL is an application for data storage, not for query purposes. To query, you must use SQL query statements to find the data in the database.
1. The Client / Server Mode Is Implemented
Microsoft SQL Server 2016 implements the client/Server mode. Client / server (C / s) computing mode is a distributed data storage, access and processing technology. SQL server is a typical example of client / server system application. This mode uses the server to store data centrally, which is convenient for maintenance and management. The client is connected to the running SQL server server through the network. The processing process of the database application is distributed on the client and the server, and the client completes the data representation and most of the business logic[15]. With the development of network technology and the continuous improvement of database transaction processing requirements, the current client / server model has developed from two-tier structure to multi-tier structure. The SQL database architecture is shown in Fig. 2.
2. Support Internet Integrated
SQL can fully support web-based XML and Internet standard database program-
ming, and send requests to access Web data and database applications through HTTP.SQL can meet the needs of large enterprises. SQL is an open database connection. It must be ensured that the system has started the TCP / IP network protocol, otherwise it cannot be implemented internet resource sharing data. In order to improve the security of data, we must encrypt the data during the development process.