Blog der Hauptbibliothek

How to find published data sets

21. November 2019 | HBZ | Keine Kommentare |

This post is also available in: Deutsch

Data should be open and accessible – to everyone and for all purposes. That is the definition of Open Data. More and more research funders are behind this, including the SNSF. On the one hand, researchers are faced with the relatively new task of making their data available for reuse. On the other hand, they themselves benefit from the accessibility of data sets, for example when writing research proposals, planning projects or discussing and comparing their own research data.

Tips for finding data sets

  • Data search in meta search engines

Meta search engines use different search engines simultaneously and can therefore deliver results from different repositories. One advantage is that users do not have to enter separate search queries in the individual repositories. A possible problem when using meta search engines is that only a limited number of repositories are searched.

OpenAIRE operates the meta search engine OpenAIRE explore, in which records from EU-funded projects can be found. Other meta search engines include datamed.org and Google dataset search, both of which are still at beta stage.

  • Data Set search in scientific repositories

Especially subject-specific scientific data repositories allow for a targeted search for data sets. To find such a repository, catalogues like re3data are available. Repositories that make it as easy as possible to re-use data can be found by limiting your search to the following criteria:

CriterionFilter Category on re3dataExample
Data should be openData Accessopen
Conditions for reuse should be describedData LicensesCC
Established Metadata Standards should be usedMetadata StandardsClimate and Forecast Metadata Conventions
Optional: Author identifiers should be enabledAID systemsORCiD

Within the repository, you can then search specifically for relevant data records. In addition to searching for repositories via re3data, we also recommend that you contact the respective research community to find out which repositories are already well established.

  •  Finding data from the data producer

In the event that the data that is of interest to you is collected by an organization, it could be that this organization offers the data in its own repository.

Federal offices and other public bodies

Opendata.swiss provides an overview of data sets made available by public authorities in Switzerland. For example, the Federal Office of Topography makes Swiss geodata available under swisstopo. The Federal Statistical Office provides statistics on a wide range of topics.

NGOs

Some NGOs operate data repositories, such as the WHO Global Health Observatory Data Repository.

Research institutions

If you are interested in records from a particular workgroup, you may also be able to find them in the appropriate institutional repository. Some universities operate an institutional data repository that contains at least some publicly accessible records. Example: ETH with its Research Collection.

Additional tip: access to data sets that are not publicly accessible

Are you interested in data that has been described in a specific publication or presented in a lecture, but that has not or only partially been shared? In his article “How to ask for data sets”, researcher Christian Kreibich mentions the following tips, among others, on how to request data directly from the researcher (Source: Christian Kreibich (April 2015): How to ask for data sets; Medium.org; quotes 13.11.2019):

  1. If a contact person is mentioned in the publication, you should write to this person. If this information is missing, contact the first author.
  2. You should use your UZH e-mail address when contacting the author and state your affiliation to your institute and your role. Moreover, if you have attended the author’s lecture, mention this.
  3. You should specify as precisely as possible the purpose for which you would like to gain access to the data.
  4. If possible, you should not ask for the entire dataset of the publication, but limit yourself to the part of the dataset you are interested in.

Abgelegt unter: Coffee LecturesOpen AccessResearch DataTips for Physicians & Health ProfessionsTips for Researchers
Tags: