Skip to Content

Wiki source code of 1 Introduction

Version 2.1 by Helena on 2025/06/07 00:40
Hide last authors
Helena 1.1 1 This document is not normative but provides a detailed view of the information model on which the normative SDMX specifications are based. Those new to the UML notation or to the concept of Data Structure Definitions may wish to read the appendixes in this document as an introductory exercise.
2
3 === 1.1 Related Documents ===
4
5 This document is one of two documents concerned with the SDMX Information Model. The complete set of documents is:
6
7 * SDMX SECTION 02 INFORMATION MODEL: UML CONCEPTUAL DESIGN (this document): This document comprises the complete definition of the information model, with the exception of the registry interfaces. It is intended for technicians wishing to understand the complete scope of the SDMX technical standards in a syntax neutral form.
8 * SDMX SECTION 05 REGISTRY SPECIFICATION: LOGICAL INTERFACES: This document provides the logical specification for the registry interfaces, including subscription/notification, registration/submission of data and metadata, and querying.
9
10 === 1.2 Modelling Technique and Diagrammatic Notes ===
11
12 The modelling technique used for the SDMX Information Model (SDMX-IM) is the Unified
13
14 Modelling Language (UML). An overview of the constructs of UML that are used in the SDMXIM can be found in the Appendix “A Short Guide to UML in the SDMX Information Model”
15
16 UML diagramming allows a class to be shown with or without the compartments for one or both of attributes and operations (sometimes called methods). In this document the operations compartment is not shown as there are no operations.
17
18 **Figure 1 Class with operations suppressed**
19
20 In some diagrams for some classes the attribute compartment is suppressed even though there may be some attributes. This is deliberate and is done to aid clarity of the diagram. The method used is:
21
22 * The attributes will always be present on the class diagram where the class is defined and its attributes and associations are defined.
23 * On other diagrams, such as inheritance diagrams, the attributes may be suppressed from the class for clarity.
24
25 **Figure 2 Class with attributes also suppressed**
26
27 Note that, in any case, attributes inherited from a super class are not shown in the sub class.
28
29 The following table structure is used in the definition of the classes, attributes, and associations.
30
31
32 |**Class**|**Feature**|**Description**
33 |ClassName| |
34 | |attributeName|
35 | |associationName|
36 | |+roleName|
37
38 The content in the “Feature” column comprises or explains one of the following structural features of the class:
39
40 * Whether it is an abstract class. Abstract classes are shown in //italic Courier// font.
41 * The superclass this class inherits from, if any.
42 * The sub classes of this class, if any.
43 * Attribute – the attributeName is shown in Courier font.
44 * Association – the associationName is shown in Courier font. If the association is derived from the association between super classes, then the format is /associationName.
45 * Role – the +roleName is shown in Courier font.
46
47 The Description column provides a short definition or explanation of the Class or Feature. UML class names may be used in the description and if so, they are presented in normal font with spaces between words. For example, the class ConceptScheme will be written as Concept Scheme.
48
49 === 1.3 Overall Functionality ===
50
51 ==== 1.3.1 Information Model Packages ====
52
53 The SDMX Information Model (SDMX-IM) is a conceptual metamodel from which syntax specific implementations are developed. The model is constructed as a set of functional packages which assist in the understanding, re-use and maintenance of the model.
54
55 In addition to this, to aid understanding each package can be considered to be in one of three conceptual layers:
56
57 the SDMX Base layer comprises fundamental building blocks which are used by the Structural Definitions layer and the Reporting and Dissemination layer
58
59 the Structural Definitions layer comprises the definition of the structural artefacts needed to support data and metadata reporting and dissemination
60
61 the Reporting and Dissemination layer comprises the definition of the data and metadata containers used for reporting and dissemination
62
63 In reality the layers have no implicit or explicit structural function as any package can make use of any construct in another package.
64
65 ==== 1.3.2 Version 1.0 ====
66
67 In version 1.0 the metamodel supported the requirements for:
68
69 Data Structure Definition including (domain) category scheme, (metadata) concept scheme, and code list
70
71 Data and related metadata reporting and dissemination
72
73 The SDMX-IM comprises a number of packages. These packages act as convenient compartments for the various sub models in the SDMX-IM. The diagram below shows the sub models of the SDMX-IM that were included in the version 1.0 specification.
74
75 [[image:1749245986724-174.png]]
76
77 **Figure 3: SDMX Information Model Version 1.0 package structure**
78
79 ==== 1.3.3 Version 2.0/2.1 ====
80
81 The version 2.0/2.1 model extends the functionality of version 1.0. principally in the area of metadata, but also in various ways to define structures to support data analysis by systems with knowledge of cube type structures such as OLAP[[^^~[1~]^^>>path:#_ftn1]] systems. The following major constructs have been added at version 2.0/2.1
82
83 Metadata structure definition
84
85 Metadata set
86
87 Hierarchical Codelist
88
89 Data and Metadata Provisioning
90
91 Process
92
93 Mapping
94
95 Constraints
96
97 Constructs supporting the Registry
98
99 Furthermore, the term Data Structure Definition replaces the term Key Family: as both of these terms are used in various communities, they are synonymous. The term Data Structure Definition is used in the model and this document.
100
101 [[image:1749245986730-349.jpeg]]
102
103 **Figure 4 SDMX Information Model Version 2.0/2.1 package structure**
104
105 Additional constructs that are specific to a registry-based scenario can be found in the Specification of Registry Interfaces. For information these are shown on the diagram below and comprise:
106
107 * Subscription and Notification
108 * Registration
109 * Discovery
110
111 Note that the data and metadata required for registry functions are not confined to the registry, and the registry also makes use of the other packages in the Information Model.
112
113 [[image:1749245986733-116.jpeg]]
114
115 **Figure 5: SDMX Information Model Version 2.0/2.1 package structure including the registry**
116
117 ==== 1.3.4 Version 3.0 ====
118
119 The version 3.0 model introduces changes in the way reference metadata are handled. In addition, it includes a few more artefacts. Finally, a few abstractions have been added, as shown in section “Basic Inheritance” in “Figure 11: Basic Inheritance from the Base Structures”.
120
121 The IM packages are largely the same.
122
123 [[image:1749245986737-968.jpeg]]
124
125 **Figure 6: SDMX Information Model version 3.0 package structure**
126
127 ==== 1.3.5 Version 3.1 ====
128
129 Whilst some additional properties have been added to Dataflow, DSD, Data Constraint, SDMX v3.1 does not change the high level information model, it remains as it is in Figure 6.
130
131
132 ----
133
134 [[~[1~]>>path:#_ftnref1]] OLAP: On line analytical processing