Process of publishing the Open Data
1. Action Plan for publishing Open Data
It is helpful to be aware that a number of steps need to be undertaken before the data is published (e.g. in Public Organizations). These include issues like:
- Mapping and understanding organizational processes (knowing what is going on, where data is produced, what the information flows are)
- Understanding the technical requirements (IT infrastructure)
- Understanding the related costs (e.g. costs required for hardware, human resources costs)
- Prioritization (what data should be published first?)
- Release schedule (what will be published when?)
It is apparent that some of these questions should be better answered by people outside of Decision Maker reach. Forming a working group, or less formally, simply involving others in a teamwork, can lead to the best results. From the experience of others, and similarly to other projects, inviting people with a “can do” mentality can be crucial.
There are several principles that might be helpful to follow:
- First publish data that is most likely to be used by anyone very soon.
- Take it one step at a time: start with a few datasets and keep on adding more.
- Don’t worry too much how that data might be useful. There may be unexpected uses, especially when this data is re-combined with data from other organizations. People will figure things out. You can feel your burden lifted because you don’t have to worry about all the uses. You don’t have to worry about managing the process of apps creation (procurement, coding, hosting, GUI design, etc.) — all of these will be handled by the users / developers who decide that your data is useful and they want to work with it.
In this post, we pointed out only a few ideas that we consider very helpful. For a much more thorough and systematic approach, see Deliverable 5.1 of project COMSODE. Deliverable 5.1 can be a basis for forming the organizations’ own Action Plan.
We recommend to work on the issues in a team because an understanding of both high-level issues known by the Decision Maker as well as understanding the technology background (known to the IT professionals) is crucial.
2. IT Professionals should co-operate with others in the organization
This part of the blog post is intended primarily for IT professionals – usually from IT department – tasked with technical execution of data publication once the decision of doing so was done. But bottom-up approach also works, so while someone within the organisation wants to approach his boss asking for particular data publication it can really ease the process if he can show how means for „technical matters“ are already found.
While in most situations IT workers are viewed by managers as responsible and competent for solving all issues of data publication, we can only recommend involve various more roles within the organisation in this process, especially:
- Data owners – they should know which data are processed, data value for organisation and for public, and should have detailed knowledge about data quality
- Legal department – they can decide if data publication is possible and to what extent, they should produce the license for data use and usually deal with formal requests for data publication from outside the organisation
- The organisation’s management – Open Data approach have several benefits for whole organisation, so it hopefully can be stated as global principle or strategy, and probably everyone knows how simple all things can be while having the director standing on your side
- IT department – in optimal situation they must solve only particular problems relating to data extraction, preparation and making data publicly accessible
Whereas there are many different ways how to solve technical tasks, many recommendations dealing with data catalogue, data schemas, ontologies, identifiers and required software architecture for Open Data publishing is covered by COMSODE Deliverable D5.1. If you intend to use Open Data Node (ODN) software created by COMSODE project, we recommend to read methodology of it’s use in Deliverable D5.2 and the User Manuals.