Note: To translate this paper into over 50 languages, please see the Automatic translation option at the top of the screen
II. Information in a statistical organization
4. There is a widespread interest across statistical organizations in being able to trace how statistical information (for example, data and metadata) "flow" through statistical business processes (into processes and out of processes). Interested parties include broad statistical systems (like the European Statistical System), National Statistical Systems (both centralized and decentralized) and smaller task teams working inside National Statistical Offices.
5. In the description of the GSIM Business group, it is seen that GSIM covers the whole statistical process and is designed to support both current and new ways of producing statistics.
6. Achieving standards-based modernization of the production of official statistics places an emphasis on being able to share and reuse processes, methods, components and data repositories. Achieving reuse of processes, methods and components will require that process designers are readily able to discover what is available for reuse and whether it may be relevant to their particular purposes and needs. The case for reuse will be challenged if, in practice, discovering potentially reusable business resources, and assessing whether those resources are actually suitable for the designer's specific purpose, takes more time than creating new design elements.
7. GSIM was designed to enable an explicit separation between the design and execution of statistical processes. The description of the GSIM Production group shows how this has been modelled.
8. There is an increasing business need to record reliable, structured information about the processes used to produce specific statistical outputs. In order to maximize transparency and reproducibility of results, it is important for a statistical organization to understand the process and its inputs and outputs. The GSIM Concepts and Structures Groups contain the conceptual and structural metadata objects that are used as inputs and outputs in a statistical business process.
9. The GSIM Base Group consists of several objects that can be seen as the fundamental building blocks that support many of the other objects and relationships in the model. These objects form the nucleus for the application of GSIM objects. They provide features which are reusable by other objects to support horizontal functionality such as identity, versioning etc. For these reasons, many of these objects are rather abstract in nature.
10. Note: GSIM information objects have been given in italics in the descriptions that follow. The diagrams included in this section are stylized representations of the model. The colours of the boxes in diagrams represent which group the information object belongs to (Blue for Business Group, Red for Production Group, Green for Concepts Group, Yellow for Structures Group and Orange for the Base Group). In many cases there is more detail to be found in the UML. Detailed information on each information object in the model, including a glossary and UML class diagrams can be found in Annexes C and D of this document.
B. Business Group
11. The Business group is used to capture the designs and plans of Statistical Programs. This includes the identification of a Statistical Need, the Acquisition, Production and Dissemination Activities that comprise the Statistical Programs and the evaluations of them.
12. An organization will react and change due to a variety of needs. In simple terms, these may be divided into at least two types of Statistical Needs: an Information Request and an Environment Change.
13. Where an organization receives an Information Request this will identify the information that a person or organization in the user community This community may include users within the organization as well as external to it. For example, a the team responsible for compiling National Accounts may need a new Statistical Activity to be initiated to produce new inputs to their compilation process. requires for a particular purpose. This request will commonly be defined in terms of a Concept or Subject Field that defines what the user wants to measure and the Population that the user wants data about.
14. When an Information Request is received it will be discussed and clarified with the user. This will be described by a Process Step. Once clarified, a search will be done to check if the data already exist. Discovering these Data Sets may be enabled by searching for Concepts and Classifications.
15. Where an organization identifies an Environment Change this indicates that there needs to be an externally motivated change. This may be specific to the organization in the form of reduced budget or new demands from stakeholders or may be a broader change such as the availability of new methodology or technology. A Statistical Need can be both internally and externally driven. For example, a statistical organization may realize that their existing Products and services must be improved. This may be in response to an Assessment of those Products and services.
Figure 1. Statistical Need
16. As shown in Figure 1, once an organization has identified a Statistical Need, it will be further specified in the form of a Change Definition. This identifies the specific nature of the change in terms of its impacts on the organization or specific Statistical Programs. This Change Definition is used as an input into a Business Case. A successful outcome will either initiate a new Statistical Program or create a new Statistical Program Design that redefines the way an existing Statistical Program is carried out.
Figure 2. Evaluation
17. At any point in the statistical business process, an organization may undertake an evaluation to determine utility or effectiveness of the business process or its inputs and outputs. An Assessment will be undertaken to evaluate any resources, processes or outputs and may refer toany object described in the model.
18. An Assessment may be of several types depending on the purpose. A Gap Analysis may be undertaken often in the context of a Business Case. An Evaluation Assessment is undertaken to determine whether a statistical output meets the need for which it was first created through analysis of:
(a) any information object that can be considered a Process Output; and
(b) in light of the original Statistical Need.
19. A Statistical Program is the overarching, ongoing activity that an organization undertakes to produce statistics (for example, a retail trade survey). Each Statistical Program includes one or more Statistical Program Cycles. The Statistical Program Cycle is a repeating activity to produce statistics at a particular point in time (for example, the retail trade survey for March 2012).
Figure 3. Statistical Programs
20. A Statistical Program (Figure 3) has an associated set of Statistical Program Designs that identify the methodology (the methods used to acquire, process and disseminate the data) used for the Statistical Program. Only one Statistical Program Design is valid for, and is identified as being used by, a particular Statistical Program Cycle. Changes to the methodology result in new Statistical Program Designs so over time each Statistical Program will have a series of designs that provide a history of changes to the Statistical Program. The Statistical Program Design identifies the set of processes that are intended to be used to undertake the activity (Process Step Design), the resources required for the processes and a description of the methodology and context.
21. Each Statistical Program Cycle consists of one or more Statistical Activities. A Statistical Activity is the set of executed processes and the actual resources required as inputs and produced as outputs. It is analogous to the Statistical Program Design but represents the execution rather than design. The same information that is identified in the Statistical Program Design and intended to be used to undertake an activity, is identified here as the actual information used. For example in the design, a dataset of a particular type may be identified as an input whereas in the Statistical Activity the filename and location of the actual input dataset would be identified.
22. The model identifies different types of activities that represent the major steps in the statistical production process (Figure 4). Three types have been specifically identified in the model but other types could be defined. The distinction between different types of activities and distinction of a Statistical Activity from a Statistical Program Cycle means that each iteration can be made up of multiple activities of the same or different types and these may or may not represent the sequence of collection through to dissemination. This model supports both the traditional approach of collecting data for a particular need, and the emerging and future approach of collecting data and producing new outputs based on existing data sources that are maintained and added to over time.
Figure 4. Statistical Activity
23. A possible future approach relates to a continuous collection process. In the age of 'big data', the cost of collecting and storing data (for example, a statistical register) is low. An organization can collect data on a continuous basis without a particular Dissemination Activity, Product or Dissemination Service in mind. In this case the organization has a Statistical Program with a Statistical Program Cycle that consists of an Acquisition Activity that gathers data and adds to a Data Resource. Any Statistical Program (consisting of only Production or Dissemination Activities) may then use this Data Resource in the future.
24. For an activity where the purpose is to acquire data a Collection Description (Figure 5) provides a description of the activity and the associated contextual information. The Acquisition Activity identifies the means by which the data is collected and where it is collected from by identifying a Data Channel.
Figure 5. Acquisition Activity
25. A Data Channel identifies the Instrument used to collect data. An Instrument is the description of the tool that will be used to collect data. Examples of it may include a questionnaire or a set of requirements to develop software for gathering data. The Instrument includes an Instrument Control and may have Question Blocks, Questions, Statements and Interviewer Instructions.
26. Once the Instrument has been designed, it must be implemented in the form of one or more Instrument Implementations. These could be printed forms, software programs, etc. The Data Channel uses the Instrument Implementation to request data and describes the technique used to do it by means of a Mode. Once the Data Channel receives the data, it sends the data to an identified Data Resource (thus populating it with Data Sets).
27. The Mode represents the way the information collection process is going to be conducted and in this way, 'how' the Data Channel is going to be used, the following table (Table 1) represents some examples of Data Channel, Instrument, Instrument Implementation and Mode.
Table 1. Examples of Data Channel, Instrument, Instrument Implementation and Mode
Data scanner device
Set of Requirements
Data Scanner Program
Web Scraping Robot
Web Service Consumer Program
Secondary transfer of data
Data Medium, File Transfer, Web Sphere Application
28. GSIM includes the notion of a Production Activity. More information about how GSIM expands on this activity can be found in the Production Group section.
29. GSIM includes the notion of a Dissemination Activity. More information about how GSIM expands on this activity can be found in the Structures Group section.
C. Production Group
30. The Production group is used to describe each step in the statistical process, with a particular focus on describing the inputs and outputs of these steps. A business process can be specified in terms of:
- The Process Steps which need to be undertaken during that process, and
- The sequence in which Process Steps need to be undertaken during that process.
31. A Statistical Activity puts into effect a statistical business process which has been designed previously (it has a Statistical Program Design) and which spans one or more phases of the business process (for example, the Collect, Process, Analyze, and/or Disseminate phases of the GSBPM).
32. At the heart of the Production Group is the description of the Process Steps within the statistical business process and the use of statistical information as inputs to, and outputs from, each Process Step. Each Process Step can be as "large scale" or "small scale" as the designer of a particular business process chooses (see Figure 6). Steps can contain "sub-steps", those "sub-steps" can contain "sub-steps" within them and so on indefinitely.
Figure 6. Process Steps can be as large or small as needed
33. In line with the GSIM design principle of separating design and production, the Production Group (see Figure 7) assumes that each Process Step will be designed during a design phase. Having divided a planned statistical business process into Process Steps, the next requirement is to specify a Process Step Design for each step. The Process Step Design identifies how each Process Step will be performed.
34. The sequencing of Process Steps within a business process is addressed through the concept of Process Control. When creating a Process Step Design, a Process Control that provides information on "what should happen next" is specified. Sometimes one Process Step will be followed by the same step under all circumstances. In such cases the Process Control simply records what Process Step comes next. However, sometimes there will be a choice of which Process Step will be executed next. In this case, the design of the Process Control will detail the set of possible "next steps" and the criteria to be applied in order to identify which Process Step(s) should be performed next.
35. During the production phase, as part of a Statistical Activity, Process Steps are executed in accordance with their design. An agent (person or system) initiates execution of the relevant Process Steps based on the following information:
- Process Step Design to determine how the current Process Step should be executed.
- Process Control to determine which Process Step to execute next.
36. A Process Step Execution Record should be recorded for each Process Step which is executed. The Process Step Execution Record is the information object which records the action. The action itself is a real world event, where Process Step Execution Record records that real world event.
Figure 7. Simplified view of Production Group objects
37. As shown in Figure 7, a Statistical Program Design is associated with a top level Process Step whose Process Step Design contains all the sub-steps and process flows required to put that statistical program into effect. Each Process Step in a statistical business process has been included to serve some purpose. This is captured as the Business Function associated with the Process Step. The Business Function, for example, might be 'impute missing values in the data'.
38. The Process Step Design associated with that Process Step will then identify the Process Method that will be used to perform the Business Function associated with the Process Step. For example, if the Business Function is 'impute missing values in the data', the Process Method might be 'nearest neighbour imputation'.
39. A Process Method specifies the method to be used, and is associated with a set of Rules to be applied. For example, any use of the Process Method 'nearest neighbour imputation' will be associated with a (parameterized) Rule for determining the 'nearest neighbour'. In that example the Rule will be mathematical (for example, based on a formula). Rules can also be logical (for example, if Condition 1 is 'false' and Condition 2 is 'false' then set the 'requires imputation' flag to 'true', else set the 'requires imputation flag' to 'false').
40. At the time the Process Step Design is executed someone or something needs to apply the designated method and rules. The Process Step Design can designate the Business Service that will implement the Process Method at the time of execution. A Business Service represents a service delivered by a person or a piece of software. Putting a publication on the statistical institute's website or putting collected response forms in a shared data source for further processing are both examples of Business Services.
41. A Process consists of a set of Process Steps, including their associated process flow information. This enables the particular set of Process Steps to be named, and potentially catalogued and reused, as a Process. Process Steps need not be grouped into named Processes unless business benefits (for example, opportunities for reuse) are likely to result from doing so.
42. A Statistical Activity initiates the execution of a top level Process Step which will result in all sub-steps being executed which are relevant to that instance of the Statistical Activity. Executing the top level Process Step should start populating a Process Step Execution Record associated with that Statistical Activity.
43. The Process Step Execution Record (see Figure 9) will record the inputs provided when executing the top level Process Step. It will then record information which allows the actual flow of execution for that instance of the Statistical Activity to be traced. This includes recording the actual inputs to, and outputs from, each sub-step as well as the evaluation of each Process Control (which, in turn, determines the specific sequence of Process Steps performed during execution).
Figure 8. Process Step Design
44. A Process Step Design (Figure 8) has a Process Input Specification that identifies the types of the Process Inputs required at the time of execution. An example might be a Process Input Specification that requires a Dimensional Dataset to be provided at the time of execution.
45. A Process Step Design may also identify Process Inputs. These refer to specific instances of inputs, rather than specifying a type of input. For example, a Process Step Design may specify that a particular Code Set will be used to provide a list of valid values.
46. Process Input Specifications and Process Inputs are often determined by the input requirements of the Business Service, Process Method and Rules associated with the Process Step Design.
47. Process Output Specifications play an analogous role to Process Input Specifications but describe the types of Process Outputs to be produced at the time of execution of the Process Step.
48. Process Control specifies what process flow should occur from one Process Step to the next at the time of execution. In some cases it may simply record the next Process Step to be executed on a fixed/constant basis. Alternatively, a Process Control may set out conditions to be evaluated at the time of execution to determine which Process Step(s) to execute next.
49. An example of the latter might be testing a Process Output against a quality criterion and initiating one course of action if the output meets the standard and another if it does not. It is not until the time of execution of the Process Step that it is possible to determine whether the standard has been met or not.
50. The specification and evaluation of conditional Process Controls refer to Rules. In the case of Process Controls, the Rules guide the process flow. (In the case of Process Step Designs, Rules guide the work done by the Process Step to produce Process Outputs).
Figure 9. Process Step Execution
51. A Process Step Execution Record (Figure 9) records the execution of activities according to a Process Step Design.
52. Execution of a Process Step uses Process Inputs in accordance with the Process Input Specification specified in the Process Step Design (Figure 8).
53. When execution takes place a particular instance of a Dimensional Dataset (for example, "Turnover of retail trade establishments by employment size, industry class and state, for November 2012") will be provided as a Process Input. The identity (instance) of the particular Dimensional Dataset may be different for each example of execution. The specific Process Inputs associated with an instance of executing a Process Step are recorded in the Process Step Execution Record.
54. Parameter Inputs are a form of Process Input used to specify which configuration should be used for a specific execution of a Process Step. For example, a set of parameters like the statistical period concerned or a sample size.
55. A Process Input may be provided to a Process Step in order for the Process Step to 'add value' to that input by producing an output which represents a transformed version of the input. Such a Process Input is classed as a Transformable Input. Usually this represents the main dataflow within the statistical process (like microdata, aggregated data, and disseminated data). It is, in short, the data transformed by the statistical process.
56. A Process Support Input influences the work performed by the Process Step, and therefore influences its outcome, but does not correspond to a Parameter Input or a Transformable Input. Examples could include:
- A Code List which will be used to check whether the codes recorded in one dimension of a dataset are valid.
- An auxiliary Data Set which will influence imputation for, or editing of, a primary dataset which has been submitted to the process step as the Transformable Input.
57. A Process Output is any instance of an information object which is produced by a Process Step as a result of its execution. Process Outputs are subtyped as part of the Process Output Specification.
58. A Transformed Output is the result which provides the 'reason for existence' of the Process Step. If that output were no longer required then there would be no need for the Process Step in its current form. Typically, a Transformed Output produced by a particular Process Step will either be provided as a Process Input to a subsequent Process Step or it represents the final product from a statistical business process.
59. A Process Metric records information about the execution of a Process Step. For example, how long it took to complete execution of the Process Step; or what percentage of records in the Transformable Input were updated by the Process Step to produce the Transformed Output.
60. Process Outputs associated with execution of the current Process Step may be evaluated as part of Process Control in determining which process step to execute next.
Figure 10. Conceptual and Structural information objects can be Process Inputs and Outputs
61. The execution of a Process Step will supply Process Inputs and result in Process Outputs. The specific Process Inputs and Process Outputs associated with the particular execution will be recorded in the Process Step Execution Record. Through Process Input Specification and Process Output Specification the Process Step Design defines the types of Process Inputs to be supplied, and the types of Process Outputs to be produced at the time of execution (See Figure 10). In many cases, these Process Inputs and Outputs are the conceptual and structural information objects that are described in the GSIM Concepts and Structures Groups (See Sections D and E). The same instance of an information object may perform different roles in different process steps.
D. Concepts Group
62. The GSIM Concepts Group contains sets of information objects that describe and define the terms used when talking about real-world phenomena that the statistics measure in their practical implementation.
63. The information objects in this group are used as Process Inputs and are often referred to in Products and Representations to provide information that helps users understand results.
64. At an abstract level, a Concept is defined in GSIM as 'unit of thought differentiated by characteristics'. Concepts are used in these situations:
(a) As a Population. To describe the set of objects it is wanted to obtain information about in a statistical survey. For example, the Population of adults in Netherlands.
(b) As a characteristic. A particular Concept about a Population is described by a Variable. The data are linked to a concept via a variable. For example, the Concept of gender in the Population of adults in Netherlands is collected by a Variable. At the representation level, there are data with Codes.
(c) As a Category to further define details about a Concept. For example, Male and Female for the Concept of Gender. Codes are linked to a Category via a Classification Scheme, for use within a Classification.
Figure 11. Populations and Units
65. As part of a Statistical Activity there is a Population (see Figure 11). There are several kinds of Populations: Target, Survey, Frame, and Analysis. The objects of interest are Units (for example, persons or businesses). Data are collected about Units. There are two kinds of Unit specified in the model. These are Observation Unit and Analysis Unit. A Unit is associated with a Population.
66. When used as part of a Statistical Activity, a Population is associated with a characteristic. The association of Population and a Concept playing the role of a characteristic is called a Variable (see Figure 11). For example, if the Population is adults in Netherlands, then a relevant Variable might be educational attainment.
67. Variable (educational attainment of adults in Netherlands) does not include any information on how the resulting value may be represented. This information is in the Represented Variable. This distinction prevents the duplication of Variable information when what is being measured is the same but it is represented in a different manner. It promotes the reuse of a Variable definition.
68. A derived Variable is created by a Process Step that applies a Process Method to one or more Transformable Inputs (Variables). The transformed Output of the Process Step is the derived Variable. In GSIM, this is modelled in the Production Group (see Section C).
Figure 12. Variable
69. A Conceptual Domain is associated with Variable. It has two subtypes: Described Conceptual Domain and Enumerated Conceptual Domain. An Enumerated Conceptual Domain, in combination with a Category Set contains information on the semantics of the Categories used by the Variable.
70.GSIM assists users in understanding both the meaning of the object and the concrete data-representation of the object. Accordingly, GSIM distinguishes between conceptual and representation levels in the model, to differentiate between the objects used to conceptually describe information, and those that are representational.
Figure 13. Represented Variable
71. The Represented Variable (see Figure 13) adds information that describes how the resulting values may be represented through association with a Value Domain. While Conceptual Domains are associated with a Variable, Value Domains are associated with a Represented Variable. These two domains are distinguished because GSIM wants to be able to talk about the semantic aspect (Conceptual Domain) separately to the representational aspect (Value Domain).
72. Both the Enumerated Value Domain and the Described Value Domain give information on how the Represented Variable is represented. The Enumerated Value Domain does this in combination with a Code List, while the Described Value Domain provides a definition of how to form the values, rather than explicitly listing them.
73.The Value Domain is defined by a Data Type. Data Types contain information on the allowed computations one may perform on the Datum (see Figure 15). For example, it is possible to distinguish between nominal-, ordinal-, interval-, and ratio-data as Data Types. Gender Codes lead to nominal statistical data, whereas age values lead to interval data.
74. A Unit of Measure refines the Value Domain . It is the entity by which some quantity is measured. Examples are Tonnes, Count of_, and Dollars.
Figure 14. Instance Variable
75. An Instance Variable (see Figure 14) is a particular Represented Variable associated with a collection of data (Datum). This corresponds to a column of data in a database. More particularly, the age of all the US presidents either now (if they are alive) or the age at their deaths is a column of data described by an Instance Variable, which is a combination of the Represented Variable "Age" and the Value Domain of "decimal natural numbers (in years)".
76. A Datum is defined by the measure of a Value Domain combined with the link to a Unit (for example, persons or businesses). A Datum is also associated with a Data Type and a Unit of Measure through the Value Domain.
Figure 15. Over view of Classification
77. Figure 15 provides an overview of the objects relating to Classifications. Classifications describe the Category role of a Concept.
78. A Classification is a categorization of real world objects so that they may be grouped, by like characteristics, for the purposes of measurement, for example ISIC (International Standard Industrial Classification of All Economic Activities). Classifications can be grouped into a Classification Family, such as industrial activity.
79. A Classification such as ISIC is a set of related Classification Schemes. It relates Classification Schemes that differ as Classification Versions or Classification Variants. A Classification Variant is based on a Classification Version. In a Classification Variant, the Categories of the Classification Version are split, aggregated or regrouped to provide additions or alternatives to the standard order and structure of the base Classification Version. A Classification Scheme has Categories organized into Levels determined by the hierarchy. A Level is a set of Concepts that are mutually exclusive and exhaustive, for example, section, division, group and class in ISIC rev 4.
80. A Classification Item combines the meaning, representation and additional information in order to meet the Classification criteria, for example "A - agriculture, forestry and fishing" and accompanying explanatory text such as information about what is included and excluded.
81. A Correspondence Table can be created by a Map that links a Classification Item in a Classification Scheme with a corresponding Classification Item in another Classification Scheme via the Category corresponding to both Classification Items. For example, in a table displaying the relationship between ISIC Rev.4 and the North American Industry Classification System (NAICS 2007 (US)), 0111 in ISIC Rev.4 is related to 111110 in NAICS.
Figure 16. Concept Systems
82. A Category is typically part of a Category Set, which is a subtype of Concept System. A Category Set contains one or more Category Items. A Category can be represented in a Category Set, a Code List or a Classification Scheme. A Category provides meaning to these information objects, for example "agriculture, forestry and fishing" or "female".
83. A Code List is also a type of Concept System. It is used for creating a group of Codes and their associated Categories. It can consist of one or more Code Items. A Code designates a Category providing representation to the meaning from the Category. For example in "F - female", the Code is F and the Category is Female.
E. Structures Group
84. The GSIM Structures Group contains sets of information objects that describe and define the terms used in relation to data and their structure. Like the information objects in the Concepts Group, the information objects in this group are used as Process Inputs and are often referred to in Products and Representations to provide information that helps users understand the structure of the data.
Figure 17. Data Resource
85. An Acquisition Program (see Figure 4) conducted by a statistical organization produces or supplies an Information Resource (Figure 17). In GSIM, one subtype of an Information Resource has been specified. This is the Data Resource.
86. A Data Resource is comprised of Data Sets. These Data Sets are made available as part of:
- an Acquisition Activity (that is, made available by the data providers for data acquisition or resulting from the Acquisition Activity); or
- a Dissemination Activity.
87. For a Data Resource, the Data Set is discovered and provided by means of the Data Location. The Data Location specifies from where the data can be retrieved. Either this can be a link to a specific file containing the data or to a Dissemination Service (see Figure 20) that will consume a query for the data and will return a Data Set. If the link is to a Dissemination Service then it is probable that the Dissemination Service is able to be queried for many types of data and so can provide many Data Sets. Each Data Set must be structured according to a known Data Structure (for example, a known structure for Balance of Payments, Demography, Tourism, Education etc.).
88. The Data Location is associated with a specific Provision Agreement which identifies the Data Provider and the Data Flow. Only one Data Structure can structure data relating to a Data Flow. A Data Flow can be grouped by Subject Fields (for example, National Accounts, Balance of Payments, Demography) which support data discovery.
89. It is mandatory that the Data Set is linked to a Provision Agreement to which it relates (that is, the union of the Data Provider and the Data Flow).
Figure 18. Data Set
90. A Data Set has Data Points. A Data Point is placeholder (for example, an empty cell in a table) in a Data Set for a Datum. The Datum is the value that populates that placeholder (for example, an item of factual information obtained by measurement or created by a production process). A Data Structure describes the structure of a Data Set by means of Data Structure Components (Identifier Components, Measure Components and Attribute Components). These are all Represented Variables with specific roles.
91. Data Sets come in different forms, for example as Administrative Registers, Time Series, Panel Data, or Survival Data, just to name a few. The type of a Data Set determines the set of specific attributes to be defined, the type of Data Structure required (Unit Data Structure or Dimensional Data Structure), and the methods applicable to the data.
92. For instance, an administrative register is characterized by a Unit Data Structure, with attributes such as its original purpose or the last update date of each record. It contains a record identifying variable, and can be used to define a Frame Population, to replace or complement existing surveys, or as an auxiliary input to imputation. Record matching is an example of a method specifically relevant for registers.
93. An example for a type of Data Set defined by a Dimensional Data Structure is a time series. It has specific attributes such as frequency and type of temporal aggregation and specific methods, for example, seasonal adjustment, and must contain a temporal variable.
94. Unit data and dimensional data are perspectives on data. Although not typically the case, the same set of data could be described both ways. Sometimes what is considered dimensional data by one organization (for example, a national statistical office) might be considered unit data by another (for example, Eurostat where the unit is the member state). A particular collection of data need not be considered to be intrinsically one or the other. This matter of perspective is conceptual.
Figure 19. Dimensional and Unit Data Structures
95. A Dimensional Data Structure describes the structure of a Dimensional Data Set by means of Dimensional Identifier Components, Dimensional Measure Components and Dimensional Attribute Components. These are all Represented Variables with specific roles.
96. The combination of dimensions contained in a Dimensional Data Structure creates a key or identifier of the measured values. For instance, country, indicator, measurement unit, frequency, and time dimensions together identify the cells in a cross-country time series with multiple indicators (for example, gross domestic product, gross domestic debt) measured in different units (for example, various currencies, percent changes) and at different frequencies (for example, annual, quarterly). The cells in such a multi-dimensional table contain the observation values.
97.A measure is the variable that provides a container for these observation values. It takes its semantics from a subset of the dimensions of the Dimensional Data Structure. In the previous example, indicator and measurement unit can be considered as those semantics-providing dimensions, whereas frequency and time are the temporal dimensions and country the geographic dimension. An example for a measure in addition to the plain 'observation value' could be 'pre-break observation value' in the case of a time series. Dimensions typically refer to Variables with coded Value Domains, measures to Variables with uncoded Value Domains.
98. A Unit Data Structure describes the structure of a Unit Data Set by means of Unit Identifier Components, Unit Measure Components and Unit Attribute Components. These are all Represented Variables with specific roles.
99. A Unit Data Structure specifies the structure of unit data. It distinguishes between the logical and physical structure of a Data Set. A Unit Data Set may contain data on more than one type of Unit, each represented by its own record type.
100. Logical Records describe the structure of such record types, independent of physical features by referring to Represented Variables that may include a unit identification (for example, household number). A Record Relationship defines source-target relations between Logical Records.
Figure 20. Dissemination Activity
101. A Dissemination Service exposes the Data Sets and other metadata that is contained in the Information Resource. It is the mechanism to create and disseminate Representations to consumers. These Representations are created dynamically on the specific request and according to the specific needs of the consumer (the Output Specification). Representations may contain any type of information, for instance statistical data (as a Data Set or visualization) or structural or conceptual metadata like a Data Structure, a Code Set or a description of a Concept.
102. A Product is the result of a Publication Activity. Products are stored for later dissemination through Dissemination Services. Examples of Products are publications, press releases, etc. Representations may be used as input to, and as components of, a Product.
F. Base Group
103.The GSIM Base Group consists of several information objects that can be seen as the fundamental building blocks that support many of the other information objects and relationships in the model. These information objects form the nucleus for the application of GSIM information objects. They provide features which are reusable by other information objects to support horizontal functionality such as identity, versioning etc. For these reasons, many of these information objects are rather abstract in nature.
Figure 21. Base Artefacts
104. The only base artefact in GSIM that gives underlying identity and naming is the Identifiable Artefact. It can be inherited by any class in GSIM for which identity, name, description, and additional documentation is required.
105. The Identifiable Artefact has three associations to Contextual String – one for each of name, description, and documentation. The value in the Contextual String is given a context by the Context Key which can be Type or Language.
106. There is no attempt in GSIM to model the administration of items in repositories such as the maintenance agency, versioning, repository functions. However, the Identifiable Artefact does have a link to Administrative Details where such details can be added using the GSIM extension methodology.
Figure 22. Organization
107. An Organization Scheme comprises Organization Items, each of which can be an Organization Unit or an Individual. The Organization Unit can be in a hierarchic scheme of Organization Items. An Individual or Organization Unit can have a number of different Contact Details.
108. The Individual or Organization Unit can play zero or more recognized roles (Organization Item Role) in the maintenance (Maintenance Agency) data collection (Data Provider) and dissemination (Data Consumer) processes.
Word version of this section only