Font Size: a A A

The Research Of Online Analytical Mining Processing (OLAM)

Posted on:2002-09-19Degree:DoctorType:Dissertation
Country:ChinaCandidate:J G CaoFull Text:PDF
GTID:1118360032455249Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
As the important technologies of Decision Support, OLAP and data mining which own their advantages and shortcomings individually while completing decision analytical tasks, but due to the reciprocity of their characteristics and functions, can cooperate and collaborate to finish the decision analytical processing. But in engineering environment, there exists some factors that cumber the seamless integration of OLAF and data mining, such as the differences of the technology backgrounds, the developing histories, the data basics, the analytical hierarchies, the customer groups. In order to lesson or discard these differences, 1. W. Han brought forward the concept of OLAM, namely OLAF Mining, in 1997. He started with the unification of the data basic of OLAF and data mining, gave the orientation that the multidimensional data model become the common data basic of OLAP and data mining, and support the applications of OLAP and data mining simultaneously. But the single data model restricts the sufficient exertion of data mining technology. This paper extends the concept of OLAM and richens the meaning of OLAM, OLAM is the abbreviation of the Online Analytical Mining Processing, which realizes the tight, seamless integration of OLAF and data mining, whose data basic includes not only the multidimensional data model, but many kinds of data models such as relation, object梤elation, object, semi梥tructure, null梥tructure data models. The research focus of OLAM is how to unit OLAM and data mining to offer efficiently the service of decision supports. This paper establishes uniform systemic theory frame of OLAM for the first time, and contributes to the system infrastructure, the conceptual model and logical model, task/transaction model, the optimizing and scheduling strategy of OLAM task, the information organization and management methods, the OLAM task define language, and the system management method. This paper puts forward 14 characteristics of OLAM, which can be the constructing rules and appraising criterions of OLAM systems. Based on these guides, the author gives the conceptual model and a kind of logical model that based on the intelligent agents. Among them, the conceptual model outlines the research scope of OLAM. This paper brings forward the work warehouse and export object store for the first time. A OLAM semantic task model which based on the constraints presents for the first time, which generalizes plenty of decision analytical algorithms, finds the common characteristics of OLAM task, extracts the elementary operations which base is the semantic slices. The definition of semantic slices that are based on the constraints, the relationship judging methods of the semantic slices are given. The relationship of the semantic slices is the basic of optimizing the OLAM transactions. A neW OLAM task/transaction model based on AOR桺AG also presents in this paper, which can describes compl,ex OLAM tasks and transactions, implies the control stream and the data stream. Based on the OLAM task/transaction model and the subtask value model that considers the main elements that confluent the importance of the subtask, many OLAM task collaboration model are given. This paper puts forward the OLPIM transaction optimizing and scheduling model, which take advantage of the strategy that the most valuable transaction preferentially are dealt with. Export objects can describe the connotation of the semantic slices, which is the key of the analytical processing of OLAM task relationships. This paper gives...
Keywords/Search Tags:OLAP, Data Mining, Decision Support, System Integration, OLAM, Transaction Model, Transaction optimization, Metadata Management, Semantic Slice, Semantics Relationship, Time Limit Guarantee
PDF Full Text Request
Related items