3 Increased data sharing and value creation
Users should not have to provide information which the public sector has already obtained. Increased data sharing is also a prerequisite for developing seamless services across sectors and administrative levels. The public sector shall share data when it can and protect data when it must. Open public data shall be made available for reuse for developing new services and value creation in the business sector.
Data is a resource that can be better exploited. All tasks performed and services developed in the public sector involve the use of data. The ways in which we exploit data is changing dramatically, and allow for completely new ways of performing tasks. Increased data sharing is also a prerequisite for developing more seamless and tailored services for users. Citizens, businesses and voluntary organisations shall encounter one digital public sector. The Digital Agenda for Norway establishes the principle that public administration shall reuse information instead of asking users to provide information they have already disclosed; the «once-only» principle.
Better provisions shall be made for reusing open data. The reuse of public information is about giving businesses, researchers and civil society access to open data from the public sector in ways that allow them to be used in new contexts, create new services and drive value creation.
Where are we?
To achieve the Digital Agenda’s once-only objective, work has begun to establish «order in one’s own house». «Order in one’s own house» has been imposed on central government agencies through the Digitalisation Circular. In order to reuse data from others for the purpose of administrative processing and performing other tasks, the agencies must first know that the data exists, where it is located and what it can be used for. «Order in one’s own house» entails good information management, and a separate guide has been prepared on this topic.9 «Order in one’s own house» is also decisive to properly safeguarding privacy. In addition to knowing where data is located and how it can be used in practice, the agencies must have a legal basis for reusing data through statutory authority or consent.
Making open public data more accessible has been a priority measure in recent years. The Freedom of Information Act gives everyone a right to reuse information that is public, under certain terms and conditions. A national data directory has been established. The National Data Directory is an overview of what data the various government agencies have registered, how it is related, and what it means. The National Data Directory marks the first step towards achieving the once-only objective.
However, the requirements in the Digitalisation Circular and the guidelines for making public data available are not adequately followed up by the central government agencies. At the turn of the year 2018/2019, only around 20 per cent of the central government agencies had published one or more data sets on www.data.norge.no, which is part of the National Data Directory. It is estimated that only around 10 per cent of relevant data sets have been made available. No special guidelines apply for municipalities concerning the reuse of data beyond the provisions in the Freedom of Information Act.
Public and private enterprises wishing to reuse public data that has not already been published on www.data.norge.no must request access to data which government agencies could proactively have made available and published as open data. Inadequate descriptions of open data from the public sector render it less visible to potential users. Moreover, Norway’s ranking in several international open data barometers is falling. The situation is better in certain areas; for example, geospatial data is published on www.geonorge.no to follow up the Spatial Data Act, and transport data is published on vegvesen.no. These data descriptions are imported into the National Data Directory.
National Data Directory
The National Data Directory is an overview of what data is registered by various government agencies, how it is related, and what it means. The data directory enables searches in the data registered by government agencies. The data directory will make it easier for government agencies and other enterprises to reuse information that is already registered. As of June 2019, the National Data Directory contained 1,259 data sets, 20 APIs, 3,326 concepts and 577 information models.10
Farm maps: https://gardskart.nibio.no
Farm maps show the land resources and area statistics for agricultural properties. They are not separate maps, but are rather composed of information from multiple sources. The solution utilises real-time data from multiple sources, and gives business owners and employees in public administration access to the same information. Farm maps support the work of farmers in connection with operational planning, documentation and applications, and at the same time enhance the efficiency of administrative processing in agricultural management. This service has been adapted for use by agricultural management authorities and by the owners and users of agricultural properties, but is open to all.
Where are we going?
The work that has been carried out in the area of information management has given us a good foundation for establishing a common overview of what data exists, what it means, and how it can be shared. In future, the value of the national data directories, the National Data Directory and data.norge.no, will depend on the agencies’ doing their share of the work on establishing the principle of establishing «order in one’s own house». This means describing their own data, concepts, information models and APIs, and actively sharing data in accordance with national guidelines. In addition, there is a need for new measures to increase the pace of the work and achieve the goals.
There is a need for enhanced competence in regulations and frameworks for data sharing and in the relationships between law and technology and between business and management models. There is also a need for more knowledge of how infrastructure in both the central and local government sectors can be adapted for data sharing. There is a need for an arena that can help data owners and users in this area and that can facilitate the exchange of experience in the public sector. Such an arena will be important in connection with developing seamless services, cross-sector digitalisation projects and work on more digitalisation-friendly regulations.
Data sharing must take place within the framework of the legal system in general, and within the scope of the privacy protection regulations in particular. Any use of personal data requires a legal basis under Article 6 of the General Data Protection Regulation. A concrete assessment of the need for regulatory amendments must be made when the need to share data arises, such as when seamless services are to be developed or where data sharing between agencies is relevant. Safeguarding the privacy of users entails, among other things, data used in a service holding the right quality and the user knowing where the information originated. Another prerequisite for data sharing is that the data is suitable for reuse; reuse must not be incompatible with the original purpose for collecting the data, and there must be conformity between the concepts used.
To avoid falling out of step with the wishes and needs of the population, the public sector must be better able to adopt the latest technologies, such as artificial intelligence and the internet of things. The new technology can be used to develop better public digital service offerings that are perceived as more relevant and adapted to the needs of individual citizens and agencies. Such adaptation will also be important for developing the jobs of tomorrow and for ensuring a more sustainable welfare society.
Increased data sharing in planning and building application processes
For private individuals and business owners wanting to submit building application to municipalities, complying with regulatory requirements for information is challenging. Both the applicant and the municipality need access to data from a number of sectoral authorities, in addition to the municipality’s own planning data, cadastral data, etc. In order to realise the benefits of a fully digital building application process, the relevant data must be adapted for digital self-service.
Through the common services platform BYGG (Build), the Norwegian Building Authority has established the infrastructure for fully digital building application processes, from completing building applications to the municipality’s decisions on building application. BYGG is a digital regulatory platform that checks and submits building applications to the relevant municipalities.
The Ministry of Local Government and Modernisation has completed a five-year programme to establish more digital planning processes, one of the objectives of which was to acquire more uniform and precise digital planning maps that conform with the planning regulations, and to develop digital planning regulations. This will result in faster and simpler building application processing for the municipalities. The Ministry will develop a digital toolkit for simpler, more uniform handling of digital land use planning matters for relevant actors.
The Norwegian Association of Local and Regional Authorities has developed a new aid for the municipalities, ePlanSak, in cooperation with the Ministry of Local Government and Modernisation and the municipalities. This is a product specification for use by the municipalities when digitalising the planning process and procuring new professional systems.
A basic prerequisite for sharing data is that we know what data is located where, what it means and whether it can be shared. A physical infrastructure for data sharing will make it easier for agencies to assess possibilities for data sharing and to facilitate the actual sharing. This work is complex for an agency to do alone. Common problems and needs can be better addressed by more centralised and coordinated performance of tasks. Some actors undertake to distribute data from one or more data providers to many data consumers. Some data providers have created, or are in the process of creating, solutions where multiple data consumers can connect. For example, the Norwegian Public Roads Administration’s Vehicle Information web service or the Norwegian Directorate of eHealth’s planned health analysis platform. For the local government sector as an administrative level, FIKS has been established as a sharing platform. There are also several sectoral sharing platforms, including PEPPOL (Difi) and Nye Feide, and a data distributor for the eBevis service in Altinn.
Norwegian Tax Administration and the a-ordning scheme
The a-ordning scheme is a coordinated service for employers to report information on income and employees to the Norwegian Labour and Welfare Administration (NAV), Statistics Norway and the Norwegian Tax Administration. The scheme is digital. Information is sent electronically, either through the employer’s payroll system or through a service in Altinn. The Norwegian Tax Administration manages the scheme on behalf of the other agencies. The intention is to simplify and coordinate reporting by employers. At the same time, there is considerable potential for reusing this information in public administrative procedures and in the private sector. The information can only be distributed to private and public actors who may receive the information with the consent of the citizen or by law. Growing demand for this data challenges the Norwegian Tax Administration’s existing core activities and places new demands on the agency’s management, uptime, security levels, etc., because other authorities and private individuals will become dependent on information that is managed by the Norwegian Tax Administration. The Norwegian Tax Administration has prepared a policy for data sharing in order to manage the growing demand.
eBevis service
The eBevis (eDocumentation) service simplifies procurement processes by allowing tenderers to avoid having to submit documentation of information which the public sector already has, such as tax certificates. When the contracting authority has a legal basis (legal authority or consent), the information may be obtained directly from the data sources, collated and delivered to public contracting authorities. The eBevis service has been developed so that the solution can also be used for other similar purposes.
In addition to sectoral sharing platforms, there is a need to consider development of a common methodology, set of principles and framework for a generic «data distributor». Such a generic methodology can be used for specific data sharing services. A central data distributor will be able to distribute data from one or more data providers to data consumers with the same needs, with regulated roles and responsibilities between the parties sharing the data. The a-ordning scheme contains information which several municipal services need. However, it is necessary to clarify the legal basis for such reuse of information, and to conduct a semantic review of the data so that the right information can be obtained from the right service. For this purpose, APIs (data retrieval interfaces) must be developed that can ensure the required data quality for use in the relevant local government services. The development of a framework for such a data distribution service is the first step towards establishing specific services (platforms) for distributing data to new relevant consumers. The Norwegian Association of Local and Regional Authorities, in collaboration with its members, will be an important actor in such work.
An ever increasing volume of information is being produced. The potential societal benefits from using big data are considerable. If conditions are made favourable, exploitation of big data can benefit many sectors, provide opportunities for enhancing the efficiency of the public sector in general and further development of services that are better tailored to users’ needs.11 Future requirements and opportunities cannot be met or exploited using yesterday’s methods and tools. A data lake is a method of storing all kinds of data, and can be likened to a central data warehouse for all types of data: structured and unstructured, documents and logs, images, audio and video. A data lake will serve as a source of all data within a given subject area that can be accessed by multiple users, and as a tool for streamlining, learning, planning and exploring opportunities, and will form an important basis for machine learning and artificial intelligence. A data lake can be built internally within an agency or as a common solution for multiple agencies or an entire sector. A data lake can facilitate efficient and standardised data sharing with secure access mechanisms.
Some examples of data lakes already exist where data can be collected and analysed by means of, for example, artificial intelligence. Such data lakes have been established, or are under establishment, by the Norwegian Directorate of eHealth (health analysis platform) and the City of Bergen (Lungegårdsvannet).
The Agency for Public Management and eGovernment (Difi) has completed a concept study that describes the current situation and future ambitions for sharing public data. The establishment of a data distributor, data lake and national resource centre for data sharing are the key recommendations from the evaluation.12
City of Bergen’s data lake: Lungegårdsvannet
In 2018, the City of Bergen established a data lake that will contain data from many different subject areas and agencies in the city and that can be shared with other parts of the city and the business sector. An initiative has been taken to establish cooperation to enable the local government sector to build on and further develop the experiences gained from Bergen. This will be viewed in connection with central government activities in this area.
Work related to information management needs to be systematised and harmonised to provide a better overview of registers and basic data, and of what data can be shared. High-quality data is a prerequisite for both complying with privacy protection legislation and making full use of machine learning and artificial intelligence.
Artificial intelligence (AI) will enable us to gain new knowledge from the large volumes of information we have in the public sector, and help us perform tasks in new ways. However, use of artificial intelligence in the public sector also raises complex questions about transparency, responsibility, due process and privacy.
Norwegian State Educational Loan Fund and artificial intelligence
The Norwegian State Educational Loan Fund has had positive experiences of using of machine learning to select candidates for «residential verification», i.e. verifying the residential address of customers registered as living away from home by checking their address against that of their parents. The Norwegian State Educational Loan Fund’s residential verification in 2018 encompassed 25,000 students. A total of 15,000 was selected based on artificial intelligence (machine learning), while 10,000 were selected in the ordinary manner by random selection (control group). The results show that artificial intelligence makes it easier to identify students who did not in fact live away from home. A total of 5.5 per cent in the control group and 11.6 per cent in the machine learning group failed the residential verification. This means that the Norwegian State Educational Loan Fund can avoid having to verify those who, with high probability, do in fact live away from home and are thus entitled to have their loan converted to an educational grant. This means that fewer students need to submit documentation, which in turn means less administration by the Norwegian State Educational Loan Fund. The students who were selected had to document that they lived away from their parents in 2017, as they had declared to the Norwegian State Educational Loan Fund.
The Government has decided to create a national strategy for artificial intelligence. This strategy will, among other things, look at:
- how we should organise the education and research sector in order to build advanced expertise in AI in Norway
- how we can make use of AI in the public sector
- how development of AI can affect working life
- the need for basic digital competence, as well as continuing and further education
- how Norway’s business sector can exploit the commercial potential of AI
Important prerequisites for enabling Norway to exploit AI are infrastructure in the form of broadband and 5G (fifth generation mobile network), computing power and, not least, collecting and structuring data and making it available
Language technology
Language technology products and services in Norwegian are a fundamental component in digitalising the public sector. Språkbanken is a service provided by the National Library offering text and voice resources for use in developing language technology in Norwegian. Common to many language technology solutions is the large volume of new and subject-specific data needed to function optimally. Securing deposits of resources to Språkbanken is a challenge.
The Government will:
- Establish a national resource centre for sharing data, with expertise in the relationship between law, technology, business and administrative processes as a learning environment and a knowledge repository for the entire public sector
- n cooperation with the Norwegian Association of Local and Regional Authorities, consider the use of existing data lakes (collections of large volumes of data), including regional and agency data lakes that can support data analysis and service development
- In cooperation with the Norwegian Association of Local and Regional Authorities, consider a generic data distributor, which entails, among other things, knowledge and methodology for how data can be properly shared through establishing principles for responsibilities, cost reimbursement and realisation of benefits
- Consider a possible obligation to publish open public data
- Prepare a national strategy for artificial intelligence