Data Approval and Auditing Process in Define
Ensures the quality of your information
To ensure interoperability between software systems, it is of paramount importance that Data Templates and their elements are clearly and uniformly described. Therefore, we have put in place a number of functionalities and processes in Define that help users create quality content. These functionalities and processes are aligned with relevant standards like EN ISO 23386, EN ISO 23387 and PAS 14191.
To ensure high-quality data across all users and to guarantee our Data Templates reflect the reality of the construction process in the real world, Define is following the processes of both Content Approval and Data auditing. While the Content Approval is an EN ISO 23386 compliant process where Domain experts judge the relevance of new concepts proposed by other users, the Data auditing process is an internal process, where Define’s Data auditors ensure consistent structuring and overall health of Define’s database. In this document, we will focus mainly on the Data auditing process but will also refer to some of the main aspects of the Content Approval process.
| Term | Definition |
|---|---|
| Data Auditing process | The process of analysis of new and existing concepts in Cobuilder Define, by which it is ensured that they conform to an “ideal state” (check Data structuring guidelines) and avoids duplication of concepts. The data auditing is not concerned whether the content is factually true or not but whether it is represented in consistent, clear and properly structured way. |
| Data auditor | User who manages the Data auditing process. Not responsible for the factual truthfulness of the concepts, but for their consistent structuring and the overall health of Cobuilder Define’s database. Only Cobuilder employees can be Data auditors. |
| Content approval process | EN ISO 23386 compliant process where Domain experts judge the relevance on new concepts proposed by other users and approve whether the concepts can be written in Cobuilder Define. |
| Domain expert | User who manages the approval process. They have the necessary competence to judge whether a new concept is factually true and relevant to the use case. Although not explicitly their job, it is strongly advisable Domain experts also check for duplicates. Cobuilder Define clients can specify their own Domain expert users, but Cobuilder can provide such service if needed. |
What is Data Auditing in Define?
All users of Define are responsible for maintaining the quality of their database in the system. But in addition to this, the Define experts take further measures to support the overall quality of the data related to the platform. This is done by executing an in-house process called Data Auditing. This is a mandatory step for users of Define and it is meant to guarantee that all users rely on the same quality of Data Templates.
Data Quality Measure 1: Prevention of duplicates
One of the major goals of the Data auditing process is to prevent the occurrence of concepts that represent the same piece of information e.g., several properties that represent the performance characteristic “resistance to fire”.
Such concepts are called duplicates and must be avoided because they:
– result in additional effort for all Define Users;
– artificially inflate the Define database;
– interfere with interoperability, since two users may want to exchange the same information but use different concepts to represent it;
Take a look at example 1, where all three properties below have similar full names, making them potential duplicates in need of deliberation:
In the case of a confirmed duplicate in Define, additional actions by Data auditors will follow like deactivation of the duplicate, its replacement with the correct concept or proposal on how to be edited so that all concepts in Define stay unique.
Data Quality Measure 2: Data Consistency
Another major goal of Data auditing is to ensure a clear and consistent representation of all concepts across all use cases.
As there are multiple Content experts (users who create new draft concepts) working independently in Define, this sometimes might result in incomplete or erroneous data input, leading to unclear or inconsistent concepts. Another factor is the evolving nature of the industry resulting in older concepts being inconsistent with newer ones.
Besides avoiding duplicates, users are responsible for creating consistent content and data auditors are responsible for the prevention of inconsistencies.
Approval Process Workflow
By signing a contract agreement, clients agree that any concepts they wish to create in Define will be subject to Data auditing in order to guarantee the quality of their data. A Domain expert must examine any new concept drafts before they become official concepts. This process ensures that concepts are relevant to the construction industry and the particular use case they are created for.
In the graph below you can find the Approval process workflow with the respective concept statuses.
Concepts that do not require data auditing are Groups of properties and Data Templates. They can be approved/rejected directly by the Domain expert without having to wait for evaluation from a Data Auditor.
The steps of the data approval process are the following:
1. Content expert completes a draft concept and sends it to be approved by a Domain expert.
2. The Domain expert reviews the proposal and judges whether the new concept is factually true and relevant to the use case according to their field of expertise (e.g. structural design, environmental evaluation, thermal analysis etc.).
The number of experts that need to evaluate a concept when it has been created is by default 1 – Domain expert. Nevertheless, this is a setting and could be changed depending on the specific needs.
If а non-auditable concept (e.g., Group of Properties or Data Template) needs to be created, it is enough for only one Domain expert to approve it. However, if an auditable concept (e.g., Property) needs to be created, in order to get to status “Approved”, it has to be evaluated by 1 Domain expert and 1 Data auditor.
When the concept is created and has been passed on for review, it will have the following “pending” status:
· Pending evaluation – by the Domain expert
3. The Domain expert rejects or approves the proposal:
- a. If the Domain expert rejects the proposal, he returns it requesting more information from the Content expert or asking him to make the necessary modifications. This feedback could be given through the comments functionality in Define.
- b. If the Domain expert approves the Draft, the proposal becomes a concept and gets published in Define with the new “approved” status (In case of a non-auditable concept).
- i. If it concerns an auditable concept, it gets passed to the Data Auditor for review
Data Auditing Process
As previously mentioned, the Data Auditing process is in place to ensure the quality of data across all users and Define clients agree that any concepts they wish to create in Define will be subject to Data auditing, before officially recorded in Define.
Concepts that require data auditing are all except Groups of Properties and Data Templates. All other concept types will need to go through the auding process after the approval by the Domain expert.
In the following workflow, you can track a typical Data auditing process from its initiation when a Content expert completes a draft concept and sends it to be approved.
The number of experts that need to evaluate a concept after it has been created are 2 – one Domain expert and one Data Auditor.
Therefore, there are two types of evaluations, which need to take place before the concept proposal is approved. When the concept is created and has been passed on for review, it can have the following “pending” status:
- Pending evaluation – by the Domain expert
- Pending audit – by the Data Auditor
The steps of the data auditing process are the following:
1. The Content expert completes a draft concept and sends it to be approved by a Domain expert.
2. The Domain expert rejects or approves the proposal:
- a. If the Domain expert rejects the proposal, he returns it requesting more information from the Content expert or asking him to make the necessary modifications. This feedback could be given through the comments functionality in Define.
- b. If the Domain expert approves the Draft, the proposal gets sent to the Data Auditor for review.
3. The Data auditor checks each draft for inconsistencies and compares it against the existing database content, whether any confirmed duplicates exist. He examines any new concept drafts before they become official concepts.
NOTE: Data Auditors are Cobuilder employees. They are a necessary part of the process, which should not be skipped, because they examine the new concept proposals, based on standards and regulations, check for duplicates with other contexts and detect inconsistencies, which are reported and addressed.
A user who has a Data Auditor role can log in the respective context and check if he has any requests for auditing by going to the left hand-side navigation, where the “Requests” are located. The Data Auditor can then select the “Audit requests” as shown below:
When the Data auditor visits the “Audit requests”, all concepts for auditing, which have been previously approved by the Domain expert, are visible there:
Also, on the right hand-side, the Data auditor is able to click on the blue balloon and open the comments that the Content creator and/or Domain expert have left:
The status of the concept at this time is “Pending Audit”. This status is used when a concept is currently waiting for evaluation by the Data auditor.
1. The Data auditor rejects or approves the proposal:
- a. If the Data auditor detects any inconsistencies or duplicates, he returns the proposal to the Content expert, suggesting how to be fixed (with the “return” button below).
If the proposal follows consistency guidelines, the Data auditor sends the draft concept for expert approval to the Domain expert (with the “approve” button below).
After the concept is approved by the Data Auditor, it will be published in Define.
Process for reporting duplicates or inconsistencies
Also, regular users are always welcome to contribute to the Data auditing process. In case users find any duplicates or inconsistencies, they can reach out to our Support team.
Then, the Data auditor analyses a report, generated by the Support team, with the duplicate or inconsistency described and:
- If no issues are found, the Data auditor returns a statement to the user
- If issues are found, the Data auditor updates the item accordingly.
