6 years ago

Supervizor is an online visualization ( of institutions' invoices. It loads data from individual open data catalogues including details on supplier, amount, expense description and budget categorization and then groups these invoices by the budget categorization and supplier in a graphic visualization making it very simple for wider public, NGOs, companies and public administration itself to gather information about on what institutions spend public money and what suppliers are they sending the money to. For example, in one click anyone can see, who are the main institution’s suppliers in consulting category, and on another click, what is the institution buying from them.

Our solution allows for other institutions to join and to visualize data in our application instance as well as copy the entire solution’s source code, adapt it to their needs and run it as their own instance on their own web server.

Our solution is based solely on web technologies allowing it to run on a generic web hosting. It was our aim, which we fulfilled, to develop the solution to run in an open-source only environment like LAMP (Linux, Apache, MySQL, PHP) as well as on Windows machine (WAMP), which is our case, and to be provided entirely as open-source itself, which we fulfilled except Google Analytics tracking code and Facebook and Twitter sharing buttons.

The original solution used our own code developed under GNU General Public License v3 plus several open-source libraries. As we put the solution on GitHub for others to use and contribute, Nette framework has been added and some code rewritten, however the open-source nature of the solution has been untouched.

The architecture consists of three parts:

1) The import manager, which is designed to be launched automatically, scans open-data catalogues of institutions involved for possible updates of their invoice dataset resources. If it finds a new version of the resource it starts an import using institution specific mapping script. The per-institution mapping script policy allows us to gather data from different formats, structures and catalogue APIs.

2) Server side MySQL database stores the whole import and groups data by institution, invoice ID, budget category, and supplier ID. REST-like API provides various queries for the client side visualization and caching of the requested data.

3) Client side draws a dynamic visualization from the requested data. The entry page consists of main budget category groups arranged into bubbles whose sizes depend on the amount of money spent in each category. After clicking one of bubbles a lucid interactive table opens with detailed filtering by budget categories and date. In the table, data are arranged by suppliers who are sorted by those with largest financial exposure towards institution as first. After clicking a supplier, list of invoices show up with description of spending, with possibility to share the invoice on Facebook or Twitter.

The solution allows for including more institutions in one instance as well as implementation several instances of the application independently.

Currently eight institutions are included in the original instance run by the Ministry of Finance of the Czech Republic: Czech Telecommunication Office, Ministry of Transport of the Czech Republic, Ministry of Finance of the Czech Republic, Ministry of Defence of the Czech Republic, Ministry of Regional Development of the Czech Republic, Ministry of Justice of the Czech Republic, Ministry of the Environment of the Czech Republic, The State Fund for Transport Infrastructure.

One institution, Prague Institute of Planning and Development, set up their own instance of the application and several municipalities expressed their interest in setting it up in close future.

The solution reuses APIs of open data catalogues used by involved institutions to gather source data. Mostly used open data catalogue solutions among the involved institutions are DKAN and CKAN software (open-source), however other solutions based on Microsoft Sharepoint and custom CMS systems are also used.

The aim of this project has always been to let other institutions join and/or reuse the software as well as to promote widespread sharing and most importantly usage of open data in the Czech Republic, therefore usage of open-source software has been one of our top priorities.

Project Supervizor is led by the open data team at the Ministry of Finance of the Czech Republic, section of Financial Management and Audit. The project includes development of the application, planning of future usage and expansion, motivation of other institutions to join and technical aid with export and open data catalogue setup. Currently development of the application source code is done at jointly at our team and using open contribution through the GitHub platform.

The project is led by one team consisting of two people with responsibilities divided as follows:

1) Project management and PR, expansion to other institutions

2) IT management, new features planning and IT support to other involved institutions

Development of application is also supported by contribution of the open-source community through the GitHub platform.

There are two big benefits of this solution for public administrations.

Firstly, Supervizor helps public institutions and municipalities to present their spending to the public in a way that everyone understands. Also Supervizor is free and open-source and therefore any public organization can use it. All that an organization needs to do is to publish the invoices according to the open data standards and ask to be included in the main application instance, or download the source code, adjust it to its needs and run your own instance.

The second and huge benefit is that Supervizor perfectly explains, is the difference between publishing information unstructured and publishing information according to the open data standards. Even though every public servant understands what invoices are, thanks to Supervizor we can easily explain what might be the benefits of their open data publication.

Until recently, transparency, especially in case public spending, was satisfied by publishing the data on organizations’ websites, either unstructured or not meeting the standards of open data in other ways. When we presented Supervizor to these organizations, they were excited about the solution and started to ask how expensive it is and what does it involve to join. We told them, that in order to join, they only need to publish their invoices according to open data standards in open data catalogue, for example one of the freeware solutions like CKAN or DKAN. Additionally we offered them consultations about project management (since we were the only big organization having experiences with managing open data project). Thanks to this initiative and thanks to Supervizor, we achieved to engage other five ministries to launch open data projects and currently we have eight organizations presenting their invoices in our Supervizor. One other organization launched their own instance of the application and lots of other organizations, mainly municipalities, are planning to launch open data projects and join Supervizor.

Supervizor is run by Ministry of Finance, but it is not based on any legislation requirement, i.e. its sustainability is not guaranteed by law and it is possible, that the next minister shuts down the Ministry of Finance’s instance of Supervizor.

However Supervizor has been intentionally developed as open-source, so that anyone else may continue to run and develop the application. The sustainability of Supervizor is therefore indirectly guaranteed by the extent of its use - if it is used by huge number of organizations, it is rather unlikely that some minister will decide to shut it down and if it happens, then participants will push towards reinstantiation of the application elsewhere.

Project is led by a team on the Ministry of Finance. Current minister of finance has provided great amount of support for publishing financial data of ministry of finance, as well as other institutions under the same political part, according to the open data standards as well as enough work-time allowance for the Open Data Project to fulfill its main goals. The project has not been provided any financial support or additional personal support as well as not extensive IT support, which has been limited to providing virtual server with basic support.           

The app is publicly located here: