In this paper,the steps of Web-based Information Mining technology are described in detail with a search instance which is designed by author firstly,and on this basis,the key technologies used in the information miming are discussed.Finally,the prospect of application of the information mining technology is put forward.

【Abstract】Based on the historical data, what-if analysis can analyze the hypothetical scenarios and provide useful information for the decision-maker. Aiming at the problem of re-computation of data cubes in what-if analysis, this paper presents an algorithm of incremental computation of data cube for Max function to reduce the access time of fact table. Experimental results show that the performance of incremental computation can be improved about 15%.

在数据库和OLAP应用中,经常需要根据历史数据以及假设数据对未来的商业行为进行预测,这就是what-if分析的过程。为了在多个决策之间进行对比,在what-if分析的过程中,需要针对相同的基本事实表进行多次假设分析,而每次假设分析都会产生一组假设数据,多次的假设分析之间也可能存在级联关系,那么对这些假设数据的存储以及处理对what-if分析具有很大的影响。文章根据聚集函数的性质将聚集函数分为可逆聚集函数和不可逆聚集函数,并针对聚集函数的类型给出了多版本what-if数据的合并规则。

Data Warehouse and On-Line Analysiscal Processing(OLAP) are the key components of Decision Support System(DSS).What-if analysis focuses on analysis on hypothetical scenarios based on historical data.It is an important type of DSS analytical processing procedure,and can provide important predicting information for DSS users.This paper firstly surveys major related work of what-if analysis and then briefly discusses the classifications from different angles.The main contribution of this paper is to present the global classification of what-if analysis.Three classical prototypes of what-if analysis based on rules,what-if analysis based on delta table and what-if analysis based on incremental CUBE maintaining are analyzed.Also the advantage and disadvantage of different type of what-if analysis and applying boundary are discussed.The authors′ recent research work of delta table based on Main-Memory Database and a new CUBE maintain algorithm are discussed to discovery new algorithms on advanced Main-Memory Database architecture.Finally the future work of what-if analysis is summarized.

The traditional model of resource sharing has the problem of insecurity of resource itself, thus, it is easy to cause malicious disclosure of shared resources. According to this, in this paper, we design a hierarchical sharing model with high security based on TPM, and describe the access mechanism of the model.

What-if analysis is an important method to analyze the hypothetical scenarios based on the historical data. It provides useful information for the decision-maker. Multiple versions are critical to what-if analysis. In this paper, we analyze the problem of multiple versions data processing and the incremental computation of cubes in what-if analysis, proposing a strategy to process multiple versions of what-if data. Our solution can adopt different data processing methods according to the type of aggregate functions, the efficiency of multiple versions data processing is improved. Furthermore, we proposes an algorithm of incremental computation of CUBE for max(min) function to reduce the access time of fact table, and the experimental results showed that the performance is improved about 10%.

What-if analysis is an important type of DSS analysis processing procedure. It analyzes hypothetical scenarios based on historical data. The data cube view must be updated when the what-if condition is changed. Since source data must be kept in order to compute the new aggregate value when new tuples are inserted or deleted, in what-if analysis, incrementally computing a data cube for holistic aggregation functions is a difficult problem. In this paper, we adopt delta cube strategy and work area technique to incrementally compute data cube for MEDIAN function. The size of work area has important influence on the efficiency of the incremental computing. This paper optimizes the size of work area based on the number and the cardinality of dimension attributes of the cuboid. Performance study shows that our algorithms are effective over large databases.

What-if analysis can provide more meaningful information than classical OLAP. Multi-scenario hypothesis based on historical data needs efficient what-if data view support. In general, delta table for what-if analysis is more general than other solutions such as query re-writing prototype of Sesame. Delta table is independent of base table and is more suitable to represent complex hypothetical updates and multi-version hypothetical updates. Due to low efficiency of traditional delta table merging algorithm which is based on set operation, there are few researches focus on delta table merging algorithm but on query re-writing algorithm. By analyzing the difference between what-if query and what-if analysis and improving the delta table merging algorithm, we propose novel algorithms without set operation of difference. Considering the feature of aggregate operations in OLAP analysis, pre-merge algorithm is presented without the generation of what-if data view before group-bys in the scenario of SUM, AVERAGE and COUNT function OLAP queries. In our experiments, the pre-merge algorithm greatly improved the efficiency of delta table merging procedure which is close to base group-by statement and superior to the re-writing algorithm. A complete comparison between all candidate delta table merging algorithms and re-writing algorithm with different what-if update conditions of update, deletion, insertion and mixed what-if updates is exhibited in our experiments, the policy of what-if analysis among different types is also discussed.

面向更新密集型应用的内存数据库系统,其检查点技术应符合几个关键的要求,包括检查点操作对正常事务处理的干扰尽可能小、能够处理存取倾斜状况、支持数据库系统的快速恢复、提供恢复过程中的系统可用性等.该文提出一种事务一致的分区检查点技术,采用基于元组的动态多版本并发控制机制,避免了读写事务的加锁冲突,提高系统吞吐能力;检查点操作以只读事务形式实现,在多版本并发控制下,避免检查点操作对正常事务处理的堵塞;由于检查点文件是事务一致的,只需要记录事务的Redo日志信息,在系统恢复过程中,只需要对日志文件进行一遍扫描处理,加快恢复过程;基于优先级的数据分区装载和恢复,使得恢复过程中新事务的数据存取请求迅速得到满足,保证了恢复过程中的系统可用性.由于采用两级版本管理机制以及动态版本共享技术,多版本管理的空间开销降低到可以接受的水平.实验结果表明,文中提出的检查点技术方案获得比模糊检查点技术高27%的系统吞吐量,同时版本管理的空间开销在可接受的范围之内,满足高性能应用的要求.

Main memory database(MMDB) has much higher performance than disk resident database(DRDB), but the architecture of hardware limits the scalability of memory capacity. In OLAP applications, comparing with data volume, main memory capacity is not big enough and it is hard to extend. In this paper, ScaMMDB prototype is proposed towards the scalability of MMDB. A multi-node structure is established to enable system to adjust total main memory capacity dynamically when new nodes enter the system or some nodes leave the system. ScaMMDB is based on open source MonetDB which is a typical column storage model MMDB, column data transmission module, column data distribution module and query execution plan re-writing module are developed directly in MonetDB. Any node in ScaMMDB can response user's requirements and SQL statements are transformed automatically into extended column operating commands including local commands and remote call commands. Operation upon certain column is pushed into the node where column is stored, current node acts as temporarily mediator to call remote commands and assembles the results of each column operations. ScaMMDB is a test bed for scalability of MMDB, it can extend to MMDB cluster, MMDB replication server, even peer-to-peer OLAP server for further applications.

In update intensive applications, main memory database systems produce large volume of log records, it is critical to write out the log records efficiently to speedup transaction processing. We propose a parallel recovery scheme based on XOR differential logging for main memory database systems in such environments. Some NVRAM is used to temporarily hold log records and decouple transaction committing from disk writes, inherited parallelism properties of differential logging are exploited to accelerate log flushing by using multiple log disks. During recovery, log records are loaded from multiple log disks and applied to data partition in time without the need of reordering according to serialization order, total recovery time is cut down. The scheme employs a data partition based consistent checkpointing method. The log records are classified according to IDs of data partitions accessed. Data partitions are recovered according to loading priorities computed from update frequencies and transaction waiting times, data access demands of new transactions coming after failure recovery are given attention immediately, thus the scheme provides system availability during recovery, which is of importance for large scale main memory database systems.

With the price of random access memory dropping and the 64 bit CPU being used widely,the techniques about Main Memory DataBase(MMDB)are studied prevailingly.And a few of prototype systems and business systems have been developed.This paper analysed the difference between the MMDB and the disk resident database in terms of query processing and index techniques,etc.This paper summarized the current research about main memory database and the open problems of MMDB.At last,it points out some directions for future research.