A posting from Stephen Katz and Marc Taconet announces the publication of a FIStatXML standard for fisheries statistical data import/export. The FAO Fisheries Global Information System (FIGIS) is a global network of integrated fisheries information; the project endeavors to provide policy makers with timely, reliable strategic information on fishery status and trends on a global scale. The FIStatXML release includes an introductory document, XML DTDs, and sample XML documents. FiStatXML "provides an intermediary format between FIGIS statistical module and reference tables, Fishstat Plus, external databases and files, reports, and other XML/HTML formats. The goal is to reduce complexity and to reuse certain components related to data import, export and transformation."

From the announcement: "This DTD is indeed fully compatible with the FIGIS dissemination and working system. It consists of four components reflecting increasing complexity in existing statistical information systems, from simplest two dimensional matrixes up to systems fully driven by Metadata such as FIGIS. This scalability feature should make it a suitable format for data exchange with any system. The DTDs have been designed to satisfy primarily Fishery statistical time series requirements, however, it should be generic enough to cover core needs that the STAT2000 project would have to cope with. This import-export format will be complemented by a "universal stats import-export application" that will assist data base managers in exporting / importing between this format and their legacy statistics data bases..."

FIGIS is "a global information system on fisheries aimed at providing policy makers with timely, reliable strategic information on fishery status and trends on a global scale. FIGIS is sponsored by FAO Project GCP/INT/715/JPN, FAO Project GCPA/INT/024/FRA, and WAICENT."

"For effective fisheries information management, FIGIS needs to promote and agree on standards: thesauri with agreed vocabularies and classifications for indexing, glossaries to ensure definitions of terms, and shared concepts. Norms for data sets content management are under development, including documentation of information quality assurance processes. FIGIS refers to the Dublin core XML Metadata standard to set up its own proposal for a Fisheries XML information standard."

FIGIS is "designed as a policy-based information system, it will enable policy makers to make informed decisions about the key challenges of sustainable development, and will support their shifting towards sustainability-centred management by providing them with a single entry point to strategic data, information, analyses and reviews of fisheries issues and trends. Integrating a galaxy of more technically specialized sub-systems, FIGIS will as well allow the user to access a variety of domains including biology, fishing technology, high seas vessels record, resources, fisheries, management systems, aquaculture, products, and markets. While organised in a data base, this information is presented in the form of fact sheets illustrated by maps, images, and statistical graphics constructed from available time series, and published via the internet, through CD-ROMs and in publications."

From the Report:

This document provides background information to FiStatXML and explain its structure. FiStatXML provides an intermediary format between FIGIS statistical module and reference tables, Fishstat Plus, external databases and files, reports, and other XML/HTML formats. The goal is to reduce complexity and to reuse certain components related to data import, export and transformation. The advantages of using a common import/export format are: (1) most of the import/export components can be reused between FIGIS, Fishstat and any future system, (2) internal import/export functionality of these systems can be made relatively simple, as it only has to process one format; (3) any new component to be developed will have access to multitude of import and export functions by other components.

A FiStatXML dataset provides everything necessary for creation of a statistical dataset from ground zero. The main four modules are: (1) Time series (TS_DATA): actual data matrix: keys, values, symbols; (2) Time series metadata (TS_METADATA): describes structure of part 1, as well as other information about the dataset; (3) Reference data (REF_DATA): what are the reference objects behind the key IDs? What are names, codes, attributes of each and how they relate to each other. (4) Reference metadata (REF_METADATA): what are types of these key values? What attributes they may have, how to sort them, how to build trees of them, etc. Additional component that is not actually a part of dataset but may in certain cases affect processing, is: (5) Processing instructions... A FiStatXML dataset is stored as an XML file or a set of files. Some of these files can be independent (a country reference table), some may require others. Some files (especially reference metadata and data parts) can be shared between datasets. Usually a complete dataset will be an XML file including a number of other files. Modules 1-4 may, and often will, be stored in separate files. For example, metadata for Capture dataset changes very rarely, reference data -- more often, time series -- even more often. So it is sensible to store these components separately and only edit/process the modified ones. Each dataset must be valid (in XML sense) according to FiStatXML DTD. This DTD must be used by anyone for validating datasets in FiStatXML..."

DTDs:

• Time Series Data Module [cache]
• Master Module [cache]
• Time Series Metadata Module [cache]
• Time Series Reference Data Module [cache]
• Time Series Reference Metadata Module [cache]
• Time Series Tag Library [cache]
• Time Series Tag Library Content [cache]

Sample data:

• Directory listing
• 'isscaap.xml' [cache]
• 'oceanarea.xml' [cache]