The Data Analysis perspective

The WebSphere® Message Broker Data Analysis perspective analyzes and filters information in XML files. You can use this analysis to create Data Analysis tools to transform your data.

XML files are flexible, but can be long and complex, or poorly structured. You might be interested in a small subset of data. However, it can be difficult to construct a model to filter and analyze this subset of data using traditional methods. Existing business intelligence (BI) tools work best with well-structured relational data.

In a Data Analysis project, you analyze a set of sample XML files according to static cues in the data. These static cues are defined in a Data Analysis profile that you specify when you create the project. When you analyze your sample documents, new data and new data structures within them are retained. Data and data structures that are not new or useful are not added. After this analysis is complete, you can explore the sample data, and select relevant elements in the Data Analysis model, which is based on the sample data that you are viewing. You decide which elements are relevant, and then add them to a target model. Use the target model to produce Data Analysis tools, which you can use at run time to transform incoming data. The Data Analysis tools that are created include maps, schema files, and subflows.

An overview of the process is shown in the following figure: The figure shows the process, as described earlier in the text.

The key steps are as follows:
  1. Create a Data Analysis project. See Creating a Data Analysis project. This action opens the Broker Data Analysis perspective, if not already open, and the Data Analysis Project Overview editor.
  2. Analyze sample XML documents. See Analyzing sample data.
    Important:
    • The sample data must conform to a schema, otherwise you cannot analyze your sample files, nor load the relevant data from within them
    • The sample data must represent your relevant data so that you can later produce a useful target model. If data is missing from your sample, subsequent input does not conform to the generated target model.
    • The size of sample documents has no limit, but the time to analyze documents increases with file size.
  3. Create a target model. See Creating a target model. Use the views in the Broker Data Analysis perspective to search for relevant information. Drag or copy nodes to the Target Model editor to construct the target model. See Populating a target model from your sample data and Modifying a target model in the Target Model editor to work with your target model.
  4. Create Data Analysis tools. See Creating Data Analysis tools.
  5. Use the Data Analysis tools to work with your data in the Broker Application Development perspective.
Avisos | Marcas registradas | Descargas | Biblioteca | Soporte | Su opinión

Copyright IBM Corporation 1999, 2013Copyright IBM Corporation 1999, 2013.

        
        Última actualización
        
        Última actualización : 2013-02-27 18:27:13


Tema de conceptoTema de concepto | Versión 8.0.0.2 | bd24240_