Design And Research Of A Tree Data Structure For Mass Data's Comprehensive Evaluation

With the rapid development of current technologies such as computer technology, network, multiple media technology, communication and so on, expert operation system which contains an amount of digital resource is operated inour various daily life. So data mining is the most important in the days when the data has been digitized generally. We must use the function of computer to mine in these amount of data. For example, sometimes government departments and large enterprises don't require or even be able to look over all of the data. But from the macro level, they need master the facts reflected from the data. Thus we need use the corresponding data mining method. There is a kindof macroscopic integrated evaluate method which adopts statistics among the various data mining methods. We define it as inegrated evaluate computer system which faces vast data, uses integrated evaluate method and uses the computer to data mining.Data mining is the process that extracts the potential valued information and knowledge used and understood from a lot of initial data which maybe fragmentary, unclear, random.Integrated evaluation refers to make the global and holistic evaluation for the object system described by the multiple attributive system structure. It consists of two problems: integrate evaluation method and evaluation index system. The former one has achieved so many research results; the latter is well associated with the specific application. Currently, for huge amounts of data, especially real-time data system ofthe index system of calculation system of the stream has reached no maturity model. Index system of computing system includes modeling, mining and index synthetic.In comprehensive evaluation of the huge amounts of data, we use Visual quantitiative numerical indicators to reflect the affection. When we analyze something, we just recognize things from a perspective of the different aspects separately in fact. So for the index value of a property, we can use several index value to describe in detail. After repeated subdivision index value, index system becomes the tree architecture. Thus tree-outline of the proposed structure of this article is designed for this kind of tree-like index system.In the background of data real-time arrival, size unknown, streaming database technology came into being at the end of 20th century, which has been applied widely in somefields, such as the financial applications, network monitoring, communication data management, web application, sensor network data processing and so on. But existing streaming data technology still cannot be easily applied in the specific area of index system of calculation. Though "Index system research based on the data warehouse" put forward an unified way to describe the index structure which put the index into the data warehouse and then mine furtherly, no concern about the related database tables, automatic construction of computing environment and configurability, straightly statistics the bottom indicator from the streaming data, index comprehensive computational model.This article references the ideas of streaming data techonology and OLAP cube analysis, based on the tree-outline of the proposed data structure, and then completes including modeling tools, indicator integrated computing software system, which can be broadly applied to various complex index system for mass data calculation, and greatly simplified index system construction of computin platform.