Data Consolidation Based On Structure Granulation And Matrix Compution

Data consolidation often occurs in our real life,such as: a number of enterprises are merged into a new enterprise,different plants are grafted into a new plant,two celestial bodies are merged into a new celestial body by collision of celestial body,many rivers are combined into a larger one,all these phenomena shows the process of data consolidation.At this time,merged data has a certain relationship with pre-merger data.How to represent data consolidation,and data association information for merging data before and after is a problem to research.This is the discussion about this article based on the following study envisaged:This process can achieve the purpose of data consolidation and retention of related information by combining data of various types in a structured way or the structure granulation.Finally,this process completes the conversion of structure and forms the theoretical method of data merging.At the same time,this theoretical approach can be programmed and used as merging model.It can be programmed by the data merging method based on matrix computing.There are calculation of matrix transformation and equivalence proof in the model.Based on this research ideas,completed the following work:(1)To clarify the concept of data consolidation,the following steps can be performed: first the merged data is defined as a kind of data set,then granulation set is used to granulate in this data set.This can forms a plan of data merging and a granulated method of merging data.(2)A weighted association structure is constructed by data set and weighted relationship,which forms the basis of data merging.The weighted granulation structure is obtained by the granulation set.The process completes the transform from weighted association structure to weighted granulation structure.Finally it established the conversion method of data merging.(3)The weighted association structure is equal to weighted association matrix,and the weighted granulation structure is equal to weighted granulation matrix.Then through the middle transformation and the target transformation,this process can realize the transformation from weighted association matrix to weighted granulation matrix.Finally it forms a matrix transformation method of data merging.(4)It is proved that the process of transforming weighted association matrix to weighted granulation matrix is equal to the conversion from weighted association structure to weighted granular structure.That is the theoretical about structural transformation can be described by the numerical calculation of the matrix.It is generally known that matrix always can be programmed,so the matrix transformation constitutes the basis of data merging algorithm.(5)This thesis combines with the practical problems in our life,explores the application of data merging method and reaches the combination of theory and practice.