Data integration apparatus, data integration method, and computer product
First Claim
Patent Images
1. A data integration apparatus collecting and integrating data in a plurality of information sources, comprising:
- a storing unit that stores meta information that includes;
a physical model, which is a model corresponding to each information source and includes data items and a primary key constraint for a data item that serves as a primary key,a logical model, which is a data model defined for a utilization-side application utilizing integrated data and includes data items and a primary key constraint for a data item that serves as a primary key, anda mapping definition that defines a correspondence relation between a data item of the physical model and a data item of the logical model as a correspondence relation between a FROM column and a TO column;
a converting unit that receives a search condition for the logical model from the utilization-side application and converts the search condition for the logical model into one or a plurality of search conditions for the physical model based on the meta information;
a collecting unit that issues the search condition for the physical model to an information source and collects from the information source physical-model-compliant data matching the search condition for the physical model;
an outputting unit that integrates data by converting, based on the meta information, the collected physical-model compliant data into logical-model-compliant data and outputs the logical-model-compliant data to the utilization-side application; and
a cleansing controlling unit that determines whether a type of the data item in the FROM column is the same as a type of the corresponding data item in the TO column, and if not, a type conversion processing unit attempts to convert the data item in the FROM column and the corresponding data item in the TO column to the same type.
0 Assignments
0 Petitions
Accused Products
Abstract
When collecting and integrating data present in separately managed information sources, data are collected from the information sources through a physical model of the information sources. Integration is executed by a process of converting a data configuration (mapping) into a logical model predefined for each utilization-side application and a process of converting and conforming values (cleansing); and the result thereof is provided as a view (logical model) for each application to the utilization-side application.
-
Citations
1 Claim
-
1. A data integration apparatus collecting and integrating data in a plurality of information sources, comprising:
-
a storing unit that stores meta information that includes; a physical model, which is a model corresponding to each information source and includes data items and a primary key constraint for a data item that serves as a primary key, a logical model, which is a data model defined for a utilization-side application utilizing integrated data and includes data items and a primary key constraint for a data item that serves as a primary key, and a mapping definition that defines a correspondence relation between a data item of the physical model and a data item of the logical model as a correspondence relation between a FROM column and a TO column; a converting unit that receives a search condition for the logical model from the utilization-side application and converts the search condition for the logical model into one or a plurality of search conditions for the physical model based on the meta information; a collecting unit that issues the search condition for the physical model to an information source and collects from the information source physical-model-compliant data matching the search condition for the physical model; an outputting unit that integrates data by converting, based on the meta information, the collected physical-model compliant data into logical-model-compliant data and outputs the logical-model-compliant data to the utilization-side application; and a cleansing controlling unit that determines whether a type of the data item in the FROM column is the same as a type of the corresponding data item in the TO column, and if not, a type conversion processing unit attempts to convert the data item in the FROM column and the corresponding data item in the TO column to the same type.
-
Specification