Statistics Croatia: use of GSBPM

In September 2012 IPA 2009 project 'Technical Assistance in Business Statistics and Quality Documentation' started in Croatian Bureau of Statistics. One of project results was a template based on Generic Statistical Business Process Model and finalized in 2014, following the GSBPM version 5.0. The template is a result of analysis carried out within the Croatian statistical system, in a way that some of individual phases of the statistical business are adjusted to the needs like collection, processing, publication and dissemination of statistical data.

The mentioned template will be used for describing, documenting every statistical survey and determining the level of quality. The template was created according to the Generic Statistical Business Process Model - GSBPM and is intended for all activities undertaken by producers of official statistics at national and international level. GSBPM is adjusted to the needs of the Croatian statistical system, designed as a model independent of the data source so that it can be used for the description and evaluation of the quality of the processes based upon surveys, censuses, administrative records and other non-statistical or combined data sources.

Metadata management was previously done at several places and metadata was not properly structured. In 2008 central metadata repository CROMETA and a tool Metadata Manager for metadata maintenance were put in production in Croatian Bureau of Statistics. Initial metadata used to fill the repository were about organizational structure, general metadata, studies and measurements. There were several sources of metadata for this initialization: Annual Implementation Plan for all statistical activities, database about publications, and so-called inventory of metadata that was done before going to developing metadata model, etc.

Central repository makes metadata available for all other systems that are parts of statistical data production. The first important consumer of metadata is Survey Processor application. This is a metadata-driven application for automatic survey processing which was one of major targets set to be developed during the project of development of Integrated Statistical Information System of CBS. That tool had to replace expensive and difficult-to-maintain mainfraime processing and to allow statisticians to be more independent from IT department.

Survey processor is not only a consumer but also a producer of metadata. Metadata is captured during survey processing which is the natural place where processing metadata is created. This metadata is immediately stored in the central repository. When a survey is processed, then following metadata is created: population, questionnaires, questions, context variables, matrix variables, matrices, registers, cubes, tables, etc. This metadata can be easily reused in to create new versions for next processing periods.

CBS is now in the final phase of migration from mainframe to Survey Processor solution. Mainframe shutdown is planned for June 30th 2015.

In 2012 CBS has developed a system for classification management KLASUS. KLASUS is the part of CROMETA and it follows Neuchâtel Terminology Model for classifications (version 2.1). KLASUS contains all of national, international and internally used classifications and code lists. Value domains which are used in processing phase (Survey Processor, other processing tools) are based on KLASUS content.

In 2013 CBS has put in production a web application for managing reference metadata about all statistical activities in Republic of Croatia. This is so-called Annual Implementation Plan application which allows statisticians to easily decribe their surveys and other activities like development projects in harmonised and standardized way. The hierarchy of themes and statistical activities follows the Statistical Requirements Compendium and is generally used as navigational structure all over the statistical production system.

IPA 2009 project 'Technical Assistance in Business Statistics and Quality Documentation'  had another important goal beside GSBPM model and that was development of Quality Database. The quality reports are based on combined structures Euro-SDMX Metadata Structure (ESMS) and ESS Quality Report Structure and are stored now in centralized Quality Database.

The accompanying tool for Quality Database is a web application which retrieves some metadata from central metadata repository CROMETA, such as descriptions of statistical activities, classifications, etc. The web application Quality Database was put in production in 2014 and enables creation of customized reports in both Croatian and English as well as of XML format files corresponding to ESMS and ESQRS structures which can be later imported to ESS Metadata Handler application. Quality reports are created during the analysis phase after the processing has finished.

Quality management is important for evaluation phase and quality control. At the moment (April 2015) quality reports are in process of preparation for several surveys and later will be published on CBS’ website.

Central metadata repository contains several metadata sections which are used globally in statistical production system. Metadata sections Organizational structure contains metadata that could be used for management process. This means that persons who are responsible for specific processes are stored as contact persons for them (not only as contact persons regarding the statistical data and publications). Contact person has a role that can be for example head of department.

Section Access and authorization rules contains metadata that provide control over the access and management of statistical processes. User group has certain privileges that allow to group members to access and run specific process on statistical data.

A web-service CROMETA is available internally for every system or tool which needs to retrieve metadata from central repository. User validation and authorization is done by one of this service's functionalities. This method retrieves information about user and all the access rights and privileges that user has for specified processes or surveys.