Advanced search
6.1.1 Prepare test data
The use and reuse of any data exchange model is greatly eased by providing adequate test data. This test data should also be made a part of the final asset uploaded to SEMIC. Thus users of your asset will be better able to understand your asset and implement it correctly. This is an auto-generated…
DISCUSSIONCreated: 21/03/2016Updated Date: 21/03/2016
6.1.2 Set quality criteria
As the owner and developer of the asset, you should define criteria against which your solution can be proofed (using the test data). This is an auto-generated Asset Development Assistant. Please refer to the library item Asset Development Assistant for further information. Category Asset…
DISCUSSIONCreated: 21/03/2016Updated Date: 21/03/2016
6.2.1 Document your Implement
For an asset to enter the clearing process on SEMIC, a comprehensive documentation is required. As described above, this will make your solution easier to maintain and extend. This is an auto-generated Asset Development Assistant. Please refer to the library item Asset Development Assistant for…
DISCUSSIONCreated: 21/03/2016Updated Date: 21/03/2016
6.2.2 Pack and upload your asset to SEMIC
Pack a zip file containing: * all artefacts relevant to your asset (e.g. XML schema, UML-models, etc.) * documentation, preferably in English * test cases and test data (if available) * the licence file This is an auto-generated Asset Development Assistant. Please refer to the library item Asset…
DISCUSSIONCreated: 21/03/2016Updated Date: 21/03/2016
6.2.3 Leave your feedback for the SEMIC community
Especially if you have reused an asset from SEMIC your feedback will be useful for any future user of this asset. So please leave your feedback detailing how you were able to use the asset, what challenges you met and what problems had to be solved. You are also welcome to suggest improvements to…
DISCUSSIONCreated: 21/03/2016Updated Date: 21/03/2016
5.1.17 Default values: Identify conflicts
Default values are often used in data sources, e.g. to signify unknown data elements. Using a date like 01/01/1970 is an example. If these default values are not adequately documented and/or different this can lead to misinterpreted data. It is therefore important to identify any default values used…
DISCUSSIONCreated: 21/03/2016Updated Date: 21/03/2016
5.1.18 Integrity constraints: Find common semantics
This type of conflict is closely related to data base design. It occurs when data from two different sources is to be integrated but is subject to different integrity constraints. Using different values as identifier keys is one example. Such conflicts usually point to a problem with semantic…
DISCUSSIONCreated: 21/03/2016Updated Date: 21/03/2016
5.1.19 Scaling and units: Identify and harmonise
Whenever an amount is stored, it is important to be aware of the unit and scale this value is used with. E.g. German brokers give the values of their stocks in Euro with two decimal places. British brokers give values in Pence with no decimals. Even if two numbers share the same semantics, they do…
DISCUSSIONCreated: 21/03/2016Updated Date: 21/03/2016
5.1.20 Scaling and units: Double-check correct use
The gravity of errors based on incorrect units and disparate scales makes testing indispensable. Prepare a set of test data and check for the accuracy of the exchanged data. This is an auto-generated Asset Development Assistant. Please refer to the library item Asset Development Assistant for…
DISCUSSIONCreated: 21/03/2016Updated Date: 21/03/2016
5.1.21 Scaling and units: Use appropriate conversion functions
If it is indispensable to maintain different units or scales of values, you must make sure that conversion is provided for accurately. This is an auto-generated Asset Development Assistant. Please refer to the library item Asset Development Assistant for further information. Category Asset…
DISCUSSIONCreated: 21/03/2016Updated Date: 21/03/2016
5.1.22 Data representations: Find agreements or conversion functions
A very common occurrence of this type of conflict is in date representation. A date given as 08/09/07 could be interpreted as 9 August 2007, 8 September 2007, or even 7 September 2008. Wherever the representation of data is not given by using a certain data type, you should look out for this type of…
DISCUSSIONCreated: 21/03/2016Updated Date: 21/03/2016
5.1.23 Aggregation: Find a common level
Data models are often radically different. This is partly due to differing requirements but to a large extent also due to personal preferences of the person modelling the data concepts. Especially the amount of detail a single concept contains varies greatly. This leads to different levels of…
DISCUSSIONCreated: 21/03/2016Updated Date: 21/03/2016