摘要
针对专家行为数据分布广、难以全面分析利用的问题,提出分布式专家行为信息系统。该系统建立安全可信的分布式数据采集体系,依据转换规则库、映射规则库、元数据对不同含义、不同格式的数据进行转换处理,形成统一的专家行为数据库,基于不同业务阶段的数据特点进行比对、关联,利用关联规则算法实现对专家行为模式的挖掘分析。结果证明该系统有助于及时、准确地得到专家行为信息。
Aiming at the problem that expert conduct data is widespread and difficult to get a complete analysis, this paper proposes a distributed expert conduct information system. This system builds secure and trustworthy distributed conduct data framework. Based on transforming rules, mapping rules, metadata references, expert conduct data with different types and meanings are transferred into a unified expert conduct database. After comparing, correlation, analyzing the data of different phrases, expert conduct patterns can be reached based on association rules mining algorithm. Results proves the system is helpful to get the expert conduct information timely and exactly.
出处
《计算机工程》
CAS
CSCD
北大核心
2009年第23期278-280,283,共4页
Computer Engineering
基金
国家科技基础条件平台基金资助项目(2005DKA33401)
关键词
分布式
数据处理
关联规则
行为分析
distributed
data processing
association rule
conduct analysis