中文—[3]高校图书馆管理的数据仓库技术研究.txt

KEJIJISNASUO ,QIKAN ,HUANYINGNIDE XIAZAI ! LIANXIFANGSHI:[e-mail]

限定为单个多维数据集;W HERE 子句用来将返回

选择数据源, 对数据仓库的数据组织进行逻辑结构
设计和物理存储结构设计, 即多维数据结构的事实
表和维表设计. 

(2) 元数据管理. 元数据是关于数据的数据, 它
描述了数据仓库的数据和环境. 通常可以分为2 类: 
一类是管理元数据, 它对原始数据及其内容、数据仓
库主题、数据转换及各种操作信息进行描述[4]; 另一
类是用户元数据, 它帮助用户查询信息、理解结果、
了解数据仓库中的数据和组织. 
(3) 图书馆决策的数据分析工具设计. 建立结构
化的决策支持查询, 实现和使用数据仓库的数据分
析工具, 包括优化查询工具、统计分析工具、C.S 工
具、OLA P 工具及数据开采工具等, 通过分析工具实
现图书馆决策支持需求. 
313　系统测试

的数据限定为特定维度和成员条件. 下面给出高校
314　系统维护
可结合多种测试方法, 对应用程序、查询工具和分析
图书管理DW 的一个典型查询: 工具等进行测试, 以确保系统的完善和健壮. 
SELECT {[Readers ]. [name ]} ON AXIS (0), 
{[Time ]. [2001]. [8], 

包括数据仓库的完备性测试和系统功能测试. 

[T im e ]. [ 2002 ]. 
图书管理数据仓库也必须进行日常维护和管


FROM Borrow
WHERE {[Readers ]. [department ]. [computer ]
}


[ 8 ]} ON A X IS (1) , 
{ [Book s ]. [ catego ry ]. [ com p u ter ]} ON 
A X IS (2) 
cube 
在上述MDX 语句中, SEL ECT 子句决定了3 
个维度, 分别是Readers 维、T im e 维和Books 维; 
FROM 子句决定析取数据时使用的是Borrow 多维
数据集;W HERE 子句将析取限制于Readers 维中
计算机系. 查询结果为2001 年和2002 年暑假期间
(8 月份) 拥有计算机类图书的计算机系学生名单. 

3　图书管理数据仓库的开发步骤

高校图书管理数据仓库的开发主要包括各种原
始数据的汇总、转储和管理, 以及利用分析工具获取
所需信息等, 具体实现步骤包括分析、设计、实施和
维护. 

311　系统分析

根据用户需求, 制定数据仓库工程的目标和计
划, 选择实现图书管理数据仓库的软硬件资源, 包括
开发平台、开发工具、终端访问工具、DBM S 以及通
信网络. 

312　系统设计

(1) 主题设计. 根据系统分析的结果确定主题, 
理, 定期进行数据更新, 特别是电子类图书的借阅数
据的更新, 以确保数据仓库正常运行. 

4　结束语

高校建立图书管理数据仓库, 就可以方便地从
大量繁杂的数据中提炼出真正有价值的信息和知
识, 帮助决策者进行决策. 一方面将推动图书馆自身
的极大发展, 同时对学校其他部门也提供重要的决
策支持. 数据仓库建设是一个复杂的过程, 高校可以
根据自己的实际情况, 选择合适的数据仓库解决方
案, 按计划稳步实施, 并在运行的过程中不断进行调
整、完善. 只有这样, 数据仓库才能真正成为图书馆
和学校的好参谋. 
参考文献: 

[1] 　张维明. 数据仓库原理与应用[M ]. 北京: 电子工业出
版社, 2002. 15217. 
[2] 　牛士英. 新时期高校情报室的前景[J ]. 情报资料工作, 
2001, (1): 122. 
[3] 　王珊. 数据仓库技术与联机分析处理[M ]. 北京: 科学
出版社, 1998. 1562170. 
[4 ] 　Inmon W H. U sing D ata W arehouse[M ]. John W iley, 
1995. 27236. 
编辑: 张新宝


. 1994-2006 China Academic Journal Electronic Publishing House. All rights reserved. http://www.cnki.net 


surface; numeric controlmachining 
L IM ing 2gui (CollegeofMechanical Engineering,Xi’anShiyouUniversity,Xi’an 710065, Shaanxi, 
China) JXSYU 2004 V. 19 N. 2p. 54257 

Meshing theoryofbeveloidgearswithcrossedaxesandcalculationof inducednormalcurvature 

Abstract:Basedon the spacemeshingtheory, itisproven thatnon2involute beveloid gearsmeshing 
with crossed axes can achieve to line contact. The meshing equation and tooth profile equation are 
presented by using meshing theory. A theoreticalway is put forward to calculate the induced normal 
curvaturealong thenormaldirectionofthecontact line. 

Keywords:crossedaxes; noninvolutebeveloidgear; linecontact; inducednormalcurvature 
L I Gui2x ian, W EN J ian2m in, LI Xiao, et al (Electromechanical college, Harerbin Institute of 
Technology, Harerbin 150001, Heilongjiang, China)JXSYU 2004 V. 19 N. 2p. 58262 

Developmentofa intelligentmonitoringandcontrollinginstrumentforoilfieldstations 

Abstract: The design of a new multi2functional intelligent monitoring and controlling instrument 
(JZK2D) foroilfieldstationsanditsimplementationarepresented. Thedesignofthehardwareisbasedon 
IntelMCS80C251 microcomputer andALTERA EPM 7128 EPLD deviceFor oilfield stations,
The LC.

. OS2II p reemp tive real2tim e m ulti2task kernel (RTOS) is applied to the design of the software. A state2machinemodelof single2wellmeasuringlogic isconstructed, anditisimplementedwithAHDLsingle2well measurement;
double2volume separator 
. JZK2D is 
excellent, and its advantages include friendly human2machine interface, powerfulmonitoring and control 
function, network support and so on. 


(Schoolof Electronic and Information Engineering, Xi’an 
JiaotongUniversity, China) JXSYU 2004 V. 19 N. 2p. 63266 
Key words: oilfield station equipment; intelligent monitoring and 
controlling instrumentchipmicrocomputer; 
LIU Wen2jiang , et al 
Xi’an 710049, Shaanxi,
; 
HU Jun, LI Jia, 

Applicationofneuralnetworktoearlydetectionof thefireofrubberbeltconveyer inmine 
Abstract: Thedetectionofthefireof rubberbeltconveyer isstudiedbymeansofaforwardneural 
network w ith three layers, and temperature, rate of temperature change, concentration of carbon 
monoxideand rateofcarbonmonoxideconcentrationchangeare taken as theparametersof thenetwork. 
Theappropriateparametersandarchitectureofnetworkareobtainedafterthenetwork istrainedwith 81 
setsofdata. Thesimulationresultshowsthattrainingtimeistheshortestanderrorreducesmostrapidly 
when theneuronsinhidden layerare tenandmomentum coefficientisequalto 0.95. The test resultof 
PVC rubberbeltconveyershowsthatwhen 0.5isregardedasthealarm thresholdandcertaintimedelayis 
employed, theneuralnetworkfiredetecting systemwillalarm asfire takesplaceabout 350 s. So itcan 
effectively implement early alarm ; at the same time, the it greatly enhances the suitability of an alarm 
system to environmental variation and anti2interference capability, and the intelligent degree of an alarm 
system is also imp roved. 
Key words: rubberbeltconveyer; fire; neuralnetwork; automaticdetection; carbonmonoxide 
GUO J ian, WANG Ru 2lin, LI N ing , et al (Department of Information and Electroni Engineering, 
ChinaUniversieyofMiningandTechnologyBeijing 100083, china)JXSYU 2004 V.19 N.2p.67270 

Studyofdatawarehouseforlibrarymanagementof collegesanduniversities 

Abstract:Agreatdealofdataisproducedinlibrarymanagementofcollegesanduniversities. Alotof 
important information for decision2making can be got by using data warehouse. Based on the newly 
research resultsof datawarehouse, the system function, constitution and star2shaped data model of the 
data warehouse for library management are discussed. A typical MDX inquiry and the development 
procedureof thedatawarehousearepresented. 

Key words: datawarehouse; star2shapedmodel; modernizationoflibrarymanagement 
ZHANG Xu 2bo, QU Zhan, LI Xiao2gang (College of Computer, Xi’an Shiyou U niversity, Xi’an 
710065, Shaanxi, China) JXSYU 2004 V. 19 N. 2p. 71273 

. 1994-2006 China Academic Journal Electronic Publishing House. All rights reserved. http://www.cnki.net