Arjan Hogenaar

Introduction

The National Academic Research and Collaboration Information System (NARCIS) is a portal giving access to all kinds of research-related information from the Netherlands. The development and maintenance of the portal is an activity of the Data Archiving and Networked Services Institute (DANS),¹ but was initiated by the Royal Netherlands Academy of Arts and Sciences (KNAW)². There is a shared responsibility for the content between DANS, the Dutch universities, the Netherlands Organisation of Scientific Research (NWO)³ and the major scientific institutions in the Netherlands. NARCIS is a free service and the budget for its maintenance and development is a regulatory task of DANS. The NARCIS portal fits well in the programme of the SURF Foundation⁴ to promote access to scientific information. Therefore frequently NARCIS has been able to profit from SURF Foundation grants. In relation to cost-effectiveness, NARCIS is being developed using open source software and standards. By doing so, vendor lock-in will be prevented and international cooperation is made easier. In the near future, the use of persistent identifiers will become more and more important. A rough indication is given of the costs involved, split into direct and indirect personnel costs, costs for hard- and software, miscellaneous costs and indirect costs.

NARCIS: a description

NARCIS⁵ is a portal giving access to a range of information sources. Started in 2006, NARCIS has now been developed into a central access point for research-related information in the Netherlands. Since February 2011, NARCIS has been a service of DANS. The most important information sources in the portal are the contents of the publication repositories of the Dutch universities and major scientific institutions and the contents of the DANS dataset repository, EASY⁶. In addition, overviews of current and completed research, Dutch researchers and their expertise and Dutch scientific institutions (together, the former Dutch Research Database, NOD) may be consulted via NARCIS. As NARCIS has become a service of DANS, a complex system of responsibilities has been evolved. DANS is responsible for the content of the components EASY and the Dutch Research Database, while the individual universities and scientific institutions are responsible for the content that has been harvested by the NARCIS portal.

Detailed information on the NARCIS portal is described in a 2006 paper,⁷ available in the repository of the Royal Netherlands Academy of Arts and Sciences (KNAW).

NARCIS is an initiative of four major players in the Dutch scientific information field: KNAW, the NWO, the Association of Universities in the Netherlands and the SURF Foundation. The rationale behind this initiative was the fact that information related to Dutch research was scattered and difficult to find. The service has been maintained by KNAW between 2006 and 2010.

DANS

DANS is an institute under the auspices of KNAW which is also supported by the NWO. Since its establishment in 2005, DANS has been storing and making research data in the arts and humanities and social sciences permanently accessible and has a specialist department for research and development that can also carry out activities for research groups.

An important activity carried out by DANS is the setting up, managing and continued improvement of the user-friendly archiving system, EASY. EASY is open to all researchers in the arts and humanities and social sciences and it allows them to permanently store their data and to search data themselves. In other words, the objective of EASY is to be the electronic repository for data from the social sciences and the arts and humanities.

Free service

NARCIS is a free service, so it has no direct revenues from its users. The model for NARCIS fits well in the mission statement of KNAW:

The Academy promotes quality in science and scholarship and strives to ensure that Dutch scholars and scientists contribute to cultural, social and economic progress.

NARCIS plays an important role in the elaboration of the goals of this mission. Therefore it was decided to develop NARCIS as a free service. Society as a whole – and the scientific community in particular – will profit from the investments in NARCIS. This makes it difficult, however, to give exact figures of costs and benefits.

There have been plans in the past to develop tailor-made paid services. The complex composition of the portal makes it difficult to decide who the owner is of specific data. To avoid possible legal problems, this idea has been abandoned.

Funding

The development and maintenance of NARCIS is funded by DANS. DANS itself is funded by the governmental organisations NWO and KNAW, though there is some extra funding. NARCIS has a collaborative partnership with the publisher SDU⁸ in the production of an annual paper version of the Dutch Research Database: Universiteiten en Onderzoeksinstellingen in Nederland. For major new developments, DANS seeks to fit with national or international programs (SURF Foundation, European Union). By doing so, in many cases 50 per cent of the development costs will be covered by grants. Several KNAW scholars have understood the importance of the funding proposal in this area and have written proposals with NARCIS as one of the main topics. So far, three proposals have been awarded. In that sense, indirectly money is being paid for NARCIS.

Sustainability

In discussing the sustainability of the NARCIS portal, one has to make a distinction between the sustainability of the service and the sustainability of the data accessible via NARCIS. The service is being maintained by DANS. DANS already plays a major role in the field of data persistency so one may assume DANS is willing to continue this service. For the data the situation is complicated. The National Library of the Netherlands is responsible for sustainable access to the open access publications described in NARCIS, whereas DANS itself is responsible for the sustainability of the datasets described in EASY (and in NARCIS). Recently PersID published an interesting report on this subject.⁹

There is one problem to overcome: there is no arrangement yet for the sustainability of the research information (the information on researchers, research institutes, projects and programmes). Within the DANS organisation, attention will be paid to this problem in future years. It is foreseen that it will be a hard task for DANS to arrange a budget for the realisation of sustainable research information. The solution will be through the development of close collaboration with the other major players in this field (KNAW, NWO, Dutch universities).

Cost-effective decision-making

During the start-up phase, no services like NARCIS existed in most Western countries. At that time, KNAW had to make many decisions on its own. Cost-effectiveness was an important aspect from the start. It was decided to build the system on open source software to prevent vendor lock-in. Meanwhile, it was possible to compare different systems in Europe, the US and Australia. NARCIS relies on standards in use in the information field such as the protocol for metadata harvesting OAI-PMH and the metadata format DIDL/MODS. By doing so, it was possible to take advantage of international developments. For instance, the new hot topic ‘enhanced publication’ is built on the de facto standard OAI-ORE¹⁰ and the representation of records is via RDF/XML. The costs for the development and maintenance of new standards for the Netherlands alone was too high, although it is sometimes a hard job to translate typical Dutch situations into the broader standards. Decisions on further developments were based on user studies in order to adapt the service to a real existing demand. As explained before, it was, and is, the intention of DANS to fit its developments with (inter-)national granting opportunities.

Responding to future developments

Sustainable access is becoming more and more important. Portals like NARCIS can only remain valuable if there is a guarantee that resources retrieved via this service are available in the long run. The costs for long-term preservation do not directly influence the NARCIS cost model, as NARCIS is just the access point to objects that are available elsewhere. In an indirect way, NARCIS has to react to this wish regarding sustainable access. The most important development in this respect is the introduction of persistent identifiers for authors, objects and organisations.¹¹

These persistent identifiers are also crucial in the development of enhanced publications.¹² Enhanced publications are composed of traditional publications and/or datasets and/or audio/video fragments and the like.

NARCIS has already adapted its infrastructure to make it possible to harvest and refer to these complex objects. Besides, the persistent identifiers may be used to provide usage characteristics of (complex) objects. Cost models for the maintenance of persistent identifiers are becoming complicated. All partners in the Dutch scientific infrastructure will have to contribute. Furthermore – with regard to cost-effectiveness – there will need to be an investigation into the profitability of joining international standards in this field.

Resource allocation

Because NARCIS was a service of KNAW in 2010 it is difficult to give an exact resource allocation. This section will offer a rough overview.Naturally, in a service like NARCIS most costs are personnel related. Of the total budget 75 per cent of the costs are salary costs. A distinction is made between direct and indirect activities.

The direct activities are as follows:

Apart from these direct activities, indirect activities may be identified separately. We define here indirect activities as the activities of employees who are not directly involved in the production of NARCIS but are necessary for facilitating the service and for making it more efficient. Such employees are, for instance, supervisors, accountants or communication consultants.

The indirect activities are as follows:

Other costs related to the maintenance of NARCIS

In the near future it will be possible to have a good indication of these costs as well. From this preliminary overview it can be calculated that the total annual costs for NARCIS are about €500,000, of which approximately €400,000 are personnel costs and the remaining €100,000 cover the cost of hard- and software plus maintenance. The total income is dependent on the number of projects NARCIS is involved in but averages €50,000 per year.

Discussion

The starting of a national service like NARCIS is not a big problem. The costs are limited to the purchase of hard- and software and the making of arrangements with the information suppliers. For these information suppliers the costs are even more restricted, because all they have to do is to make sure their data are harvestable. The real problems start after a few years, when people have become familiar with the service and have begun to ask for additional features. Asking is simple, but realising is dependent on cooperation with the suppliers. Therefore extra time and money is needed both at the NARCIS level and at the level of universities and scientific institutions. At this time, NARCIS is on the threshold of a new era with persistent identifiers, linked data and enhanced publications: all developments that researchers have requested. These are new challenges, not only technically, but also for the cost model of the service. Implementing all these new features is too expensive for a single organisation. Therefore a portal like NARCIS needs a national approach, in which all scientific partners contribute to its maintenance and development.