《浅谈数据挖掘技术在信息系统审计中的实际运用（Application of data mining technology in information system audit）.doc》由会员分享，可在线阅读，更多相关《浅谈数据挖掘技术在信息系统审计中的实际运用（Application of data mining technology in information system audit）.doc（9页珍藏版）》请在三一文库上搜索。
1、浅谈数据挖掘技术在信息系统审计中的实际运用（Application of data mining technology in information system audit）Application of data mining technology in information system auditAbstract this paper introduces four kinds of commonly used data mining technology in the audit, and probes into the application of the technology i
2、n the information system audit, focus on the data generated by the information system itself how to conduct data mining analysis, found that the audit doubts made an exploration and Research on data mining technology in the future, how to effectively use the the analysis of development strategy in i
3、nformation system audit.Key words: audit; information system audit; data miningWith the rapid development and wide application of information technology, audit management, mode of management and accounting is undergoing significant changes, more and more dependent on the information system of high e
4、fficiency, complex and large, at the same time constitute accounting, the traditional management system to change. The raw data that auditors face is no longer just manual vouchers, books and reports, but the information system itself and a large amount of electronic data that is highly centralized.
5、 Information system audit, as a new auditing way under the information environment, is increasingly showing its important status.At present, one of the most difficult problems faced by information system audit is how to deal with the challenges of increasing data mass. The data here refers not only
6、to the financial and business data of the audited units, but also to the data closely related to the information system itself, such as the log data, the progress of the system, the cost data, etc. It is necessary to quickly and effectively discover the laws and characteristics contained in the larg
7、e amount of data, dig out audit clues, expand audit results and reduce audit risk. To achieve this goal, data mining technology has become a golden key, which provides new ideas and methods for information system audit.First, data mining overviewData mining is a process from which a large amount of
8、incomplete, noisy, fuzzy and random data is extracted, which is unknown, but is potentially useful information and knowledge. Data mining plays an important role in the process of information system audit. Because it can not only produce the information system by deep analysis and research of financ
9、ial and business data audit unit, reveal the original characteristic and internal relations, but also in the application of the algorithm to achieve its development of information system audit, security audit etc. through model matching and mining. In addition, through the effective combination of d
10、ata mining technology and technology information system audit can determine the selection point of snapshot snapshot technology, embedded audit procedures and determine the file system control what information should be collected.Two, the application of data mining technology in information system a
11、uditThere are many data mining techniques. The author only enumerates several common techniques that are most likely to be used in auditing. These technologies can be used for comprehensive analysis and processing of the audited data, so as to improve the efficiency and quality of the information sy
12、stem audit and reduce the audit risk.(I) using statistical analysis techniques to find deviation dataStatistical analysis technology refers to the scientific analysis of the function relation or correlation between database item items by using the principle of statistics. Specific methods include st
13、atistical (for large data of the maximum and minimum average value, summary, etc.), regression analysis (using the regression equation to represent the number of relationships between variables, the difference (between) analysis to determine the overall parameters from the difference is whether diff
14、erences exist in the sample statistics).This technology is a mature technique for data mining in auditing. The audit staff through the establishment of a statistical model to collect the audited units and units of the same type of a large number of financial and business historical data analysis, mi
15、ning function relationship exists within or related, then the relevant data on the audit period forecast. By comparing the predicted and audit values of the analysis, it helps auditors to spot the doubt. For example, deep excavation of certain financial indicators, you can find abnormal index data,
16、access to audit clues.Statistical analysis technology can be well used in the development and audit phase of information system. As the most important part of the development stage audit, it is the review and control of the information system project schedule and cost. The audit staff based on the e
17、xperience of the project of the same type of data, statistical analysis model based on the project schedule and cost, in accordance with the project plan to achieve the predicted schedule cost, and then the actual project cost, the cost is not to judge the effective control of. Through the departure
18、 from the larger expenditures for further review, determine the implementation of the project in the process of rational allocation of resources, there is no loss, waste and other unreasonable expenditures.(two) using cluster analysis technique to determine audit emphasesCluster analysis is to decom
19、pose or divide a data set into different groups, so that the objects in the same set are as similar as possible, and the objects in different groups are as different as possible. Through clustering, it is easy to identify dense and sparse regions, and finds the relationship between global distributi
20、on patterns and data attributes.In the process of auditing, it is usually used to cluster the same type of financial data or business data of the audited units in the information system by clustering analysis, so that they can be clustered into similar characteristics. Generally speaking, the change
21、 of financial data and important business data (such as sales data) has a certain regularity. Therefore, if some data is in a sparse area, indicating that its abnormal performance changes, the need for attention. At the same time, by observing the characteristics of the regional records, we can find
22、 the characteristics of the audit needs to be verified.In the audit of information system security, auditors can use the technology to realize the examination of computer logs. The computer log records all kinds of important events in the computer system. Through it, we can understand the system ope
23、ration, audit the security incidents, and find the error anomalies, so as to give a comprehensive evaluation of the system security. However, there are many kinds of computer logs, such as large format difference, large series growth and great job pressure for auditors. In this case, clustering anal
24、ysis plays a multiplier role. The log data of similar data into the audit staff will be collected to mining tools, such as SQL2005 Microsoft Analysis Services tool, select the clustering model, can achieve the classification of log data using its powerful analysis function, magic non normal log reco
25、rds to distinguish from log normal data. For example, in the middle of the night time login log records, log records and other violent attacks, providing clues for subsequent review of the evaluation.(three) using outlier analysis technique to excavate audit doubtful pointsOutlier refers to data tha
26、t deviates significantly from other data, i.e., that does not satisfy the general pattern or behavior. Outlier analysis is an important technique in data mining. It is used to discover abnormal data and abnormal behavior in data source which is significantly different from other data or behavior.In
27、the regular audit, the outlier data is often the auditors rely on occupation sensitive, with a pair of piercing eye to find the. In the face of electronic data, auditors need strong data by computer analysis, the outlier detection algorithm, abnormal audit data or abnormal frequency, so that there m
28、ay be hidden irregularities. In particular, the isolated point analysis method, such as the combination of information system and audit technology, can achieve more satisfactory results. To determine the outliers by using this method, the audit staff can be selected in the audit sample outlier data,
29、 through the use of information systems audit, such as the use of snapshot capture data, real-time monitoring method using embedded audit module, to achieve the purpose of locking the data.For example, in a hospital information system audit, the auditor selects all the details of the hospital outpat
30、ient prescription data base in the first half of the year, the number of days of prescription on the same patient within half a year with a drug in more than two pieces of prescriptions were interval outlier analysis. The results show that there are several distinct low values compared with the norm
31、al data distribution graph states. The isolated points are distributed over 02 days. After extensive audit, found that the hospital has a prescription drug prescribing. Therefore, auditors select the audit samples on the isolated point data, and use snapshot technology to quickly capture the doubtfu
32、l records in the annual data, which greatly improves the audit efficiency.(four) using association analysis technique to reveal key attributesAssociation analysis techniques extract frequent patterns from all the details or transactions of an operation database, and then summarize a set of events or
33、 entries that relate to other events or items.In the financial or business data audit, similar or not similar accounting items and data items, there may be some correspondence analysis method to find and analyze the correlation, can find some hidden economic activities, finding out all the attribute
34、 data between the mutual influence, to provide the reference for the follow-up audit work. Continuing with the hospital information system audit project, the audit team conducted a special analysis of inpatient costs to seek the most important factors affecting their cost formation. First of all fro
35、m the inpatient expenses table selected patient age, disease diagnosis, hospital wards, treatment results, hospitalization days, drug expenses and other properties as the analysis, and then establish the correlation mining model, by using the strong correlation principle, obtained the mining results
36、: the drug expenses and hospitalization determines the total hospitalization expenses of patients. Therefore，Auditors focused on reviewing the two important attributes and eventually found the hospitals irregularities, hospital days, rules of calculation, and the issue of over standard drugs. The te
37、chnology can also be used in conjunction with information systems auditing techniques.Three, data mining technology in the information system audit in the effective use of development strategies(1) developing efficient and easy-to-use data mining tools;At present the common data mining tools are mos
38、tly generic data mining software developed by foreign companies, most of the audit staff is difficult to achieve in the daily audit master application, which has become the main reason for restricting the data mining technology is widely used in the information system audit in one. Therefore, it is
39、urgent to develop a data mining tool that meets the needs of auditing and is easy to learn, efficient and easy to use. This set of tools can be developed as a separate mining application software, can also be integrated in the software function module of audit information system, but must cover a va
40、riety of commonly used data mining technology, and the fusion of several typical information system audit data mining model example. Only in this way can we help auditors quickly define specific business, select appropriate mining techniques, establish models accurately and improve work efficiency.(
41、two) timely introduction of information systems audit operations guide;The audit authority should be organized as soon as possible the introduction of relevant information system audit standards and audit guidelines, especially the operation rules of the information system of data mining, such as sy
42、stem log analysis, to standardize and guide the practice of information system audit, improve the quality of data mining, to prevent and avoid the audit risk.(three) establish a complete team of professionals in Information System AuditingTo be skilled in using the data mining technology, carry out
43、the information system audit, audit institutions and auditors should raise awareness, attaches great importance to establish the spirit to tackle tough. At the same time, the audit institutions at all levels should intensify efforts to carry out various forms of training, training and introduction of a number of familiar and proficient in auditing and computer expertise, master data mining technology high-end complex talents. (