XML Based Metadata
Metadata here contain two parts. XML file, and XML schema. XML files are written using a number of elements and attributes to describe various datasets. There are four types of XML files created with different functions, which are (1) To describe category of data files by data sources or by physical data types, (2) To describe storage information of data files, (3) To describe content and semantic of data files and allow system to understand the meaning of the data, and (4) To describe format and organization of data files. XML schema defines the elements and attributes used in XML files, and describes the structure of an XML document and specifies legal building blocks of an XML file, which can be used to validate an XML file. Managing Metadata with XML Editor As stated, the key to manage data is to manage their metadata. However, for scientists not familiar with XML, writing XML metadata for new data source or managing existed XML files can be potentially complicated. However, there are many free or commercial, XML editors, which provide user-friendly GUIs for scientists to specify the structure, storage, category and content of the earth science data files without knowing much of the XML schema and rules. The editor makes it convenient for users to manage metadata. Flexibility and Extensibility. By employing the XML metadata technology, our system is so designed that it is flexible to be extended or changed with any new needs and any incremental development of the system. (1) Adding New Data Sources The user can add new data sources easily by simply capturing information about the new data source and creating new metadata files to describe it with an XML editor. There is no need to change the system. (2) Creating New Applications Based on XML metadata, we have implemented data management, search and retrieval, and some applications. Currently, applications we have implemented include data viewer, analytical functions and scientific visualization methods.This will develop relevant data mining, fusion and some other analysis applications based on metadata. Scientists may also utilize existed XML metadata to develop their own analysis toolkits with our system. (3) Dynamic Applications Enabling Since there are a variety of datasets in the system, apparently, not every application is suitable for all kinds of datasets. For example, for a time-series featured dataset, 2-D (twodimensional) plots may be applicable while 3-D (three-dimensional) visualization may be not. How do users know which applications could be employed to a specific dataset? To achieve our goal of transparency to users, we have built metadata to describe corresponding relationships between dataset models and available applications. Based on these metadata, the system will configurate GUI dynamically for the dataset retrieved, so that only those relevant applications will be showed on the GUI, which facilitates users to pick up suitable applications provided conveniently when analyzing the retrieved data online.