Modeling Applications- Information / Data

Post Reply
Rick.Allen
Posts: 23
Joined: 21 Jun 2012, 18:08

In producing report,

View: Information/Data Summary for Application

All sections but the Data Managed are populated.

How do I populate this section? What is function of the 'persist' flag in this context?

Also, it seems that the Documentation section has a lot of data that is not part of the report.

Thank you. Rick
Rick Allen
Enterprise Resource Planning (ERP) Analyst
Hamilton Police Services, Ontario, Canada
User avatar
jonathan.carter
Posts: 1087
Joined: 04 Feb 2009, 15:44

Hi Rick,

Thanks for posting.

This view embodies the following concepts:

- an Application persists information (Information Representation instances) elements through its ‘uses_information_representation’ slot when the “persists information” flag is set to ‘true’ (app_pro_persists_info_rep). The key class for capturing these is the APP_PRO_TO_INFO_REP_RELATION, on which you’ll find the “persists information checkbox”
and which connects the Application to relevant Information Representations.

- in the context of this persisted information (persisted by your selected application), the set of Information Store instances that contain this information are found. This is calculated by finding those Information Stores and looking at the ‘deployment_of_information_representation’ slot for instances of Information Representations that feature in the APP_PRO_TO_INFOREP_RELATION instances from the previous step.

- Knowing the relevant Information Stores (physical layer things), we can then find the Data elements that are contained in these information stores and from there, we can find the Data Representation and Data Object elements that are going to be displayed in the table. Note that the APP_PRO_TO_INFOREP_RELATION instances are queried here to find things such as the data acquisition method and the source of this data, via ANOTHER relationship class, the APP_PRO_TO_INFOREP_TO_DATAREP_RELATION, in the “Specific Data Operated On” slot.

Let’s summarise the above View query flow in terms of what elements should be populated.

First, the APP_PRO_TO_INFOREP_RELATION (instances held in the Application Uses Information slot) hold the details of what information is being used and how the application operates on it.

Then, connect the Data elements that are used to deliver the information via the “Specific Data Operated On” slot - creating new instances. Each of these describes the Data Representation (which should be mapped to a Data Object), the Acquisition Method (how do we get this data). We can capture more details about the CRUD on this Data in the context of it being used to deliver Information X, used by Application A.

This gives us the logical application perspective and - most importantly, defines which Data is being operated on by the in-focus Application - but the View also wants to show us things like data completeness or other quality aspects that can only be captured at the physical level. This is a bit more straightforward. We need to create instances of Physical Data Object and connect these to:

- The Data Representation (one of those described above, typically)
- The Information Store in which we’d find instances of that data.

Here we can define the Data Quality attributes of this physical data element to describe completeness, organisation scope etc.

This is a bit complex but hopefully this helps to explain why the meta model works like this and how to connect up the relevant instances. I presume that you have the Application elements, the Information elements and the Data elements, in which case these need connecting up via the relationship class instances.

Jonathan
Essential Project Team
Rick.Allen
Posts: 23
Joined: 21 Jun 2012, 18:08

Jonathan,

Thank you for your reply.

I had all the basic elements in place. Once I got the correct relationships defined as per your explanation all was well.

I am still not quite clear when to Persist or leave unchecked for relationships.

Thanks. Rick
Rick Allen
Enterprise Resource Planning (ERP) Analyst
Hamilton Police Services, Ontario, Canada
User avatar
jonathan.carter
Posts: 1087
Joined: 04 Feb 2009, 15:44

Hi Rick,

That "Is persisted" checkbox is used to say that the data is being persisted - that is, the Application 'saves' the data that is selected on the relationship. The relationships capture the context in which the data is being used by the application.

If your application saves the data, e.g. to a datastore that it 'owns', then you should check the 'persisted' box.

Jonathan
Essential Project Team
Post Reply