Search Tutorial
Database search is performed using the Lucene Query Syntax, but be aware that while Lucene uses OR as the default conjunction operator, DDF uses the AND operator. So when no boolean operator is specified between search terms, AND will be used.
The available search parameters for publications are:
| Fields | Field names |
|---|---|
| Author |
au: or author: or authors:
Example: au:"Smith, Bob" |
| Editor | ed: or editor: or editors: |
| Affiliation | af: or affiliation: |
| Title | ti: or title: |
| Journal title / abbreviated title / original title | jo: or journal: or journaltitle: or journal_title: |
| Abstract | ab: or abstr: or abstract: |
| ISBN | isbn: |
| ISSN | issn: |
| DOI | doi: |
| Publication year | y: or year: |
| Publisher | pu: or pub: or publisher: |
| Volume | v: or vol: or volume: |
| Issue | i: or iss: or issue: |
| Page start | p: or page: |
| Keyword |
kw: or ke: or key: or keyword: or keywords: or su: or subj: or subject:
There are no controlled keywords in the Danish National Research Database, however, searching for keywords might return highly relevant records. |
The available search parameters for researchers are:
| Fields | Field names |
|---|---|
| Name |
name
Example: name:"Josiah Carberry" |
| ORCID |
orcid
Example: orcid:0000-0002-1825-0097 |
Stemming
Any search term is by default reduced to its elemental root in order to retrieve all forms of the term, just like applying truncation. If you search for the term test you will also retrieve hits on tests, tested, testing, tester, etc.
Note, that stemming is only working optimally in connection with english terms. It is recommended to use truncation on terms in all other languages, if needed.
Elemental roots of english terms can be found on: http://snowball.tartarus.org/algorithms/porter/diffs.txt
For more information on stemming, please refer to: http://snowball.tartarus.org/algorithms/english/stemmer.html
Truncation and wild card matching
In most cases truncation is not necessary because of the default stemming of search terms. Asterisk is used as wildcard symbol, representing any character within a word. Searching for wom*n will retrieve woman and women.
The following examples illustrate the difference between a query without truncation (query 1) and a query with truncation (query 2):
Query 1: ti:water
Retrieves hits on: water and
waters (default stemming)
Query 2: ti:water*
Retrieves hits on: water, waters
and waterlogged.
Ranking
Search results can be ranked on various parameters. The default is “relevance”, giving a priority boost to the title data element for publications and a priority boost to the name element for researchers. Relevance for researchers is also affected by their number of registered publications such that a researcher with many publications will be ranked higher than a researcher with the same name but few publications.
To refine search results, the database offers a selection of facets, which allows for narrowing down of the various parameters.
Most recent publications on researcher profile
On the researcher profile a list of up to 10 of the researcher's most recent publications is displayed. The publications are listed by publication year starting from the most recent.
Get
For publications found in the database, it provides: backlinks to the research institutions' CRIS systems; links to Open Access versions of full texts; and links to publishers' editions.
Thus, access to the full texts of publications found in the database is handled by the user interfaces of the CRIS systems in which the publications were originally registered. They may offer different and differentiated access options – based on user role, affiliation, existence of deposited Open Access versions, etc.
This database does not contain digital copies of any of the publications and has no service with regard to ordering of material. For ordering, please refer to the research institutions’ CRIS, a library and/or the publishers of the publications.
API
Search request can use content negotiation to force the response in different formats. Default is HTML.
For any URL https://forskningsdatabasen.dk/en/catalog?PARAMETERS, adding
- “.json“ (https://forskningsdatabasen.dk/en/catalog.json?PARAMETERS) will force a response in JSON format
- “.xml“ (https://forskningsdatabasen.dk/en/catalog.xml?PARAMETERS) will force a response in XML format
A JSON response contain exactly the same as an HTML response: a number of matching records and facets (including facet values and counts). An XML response only contain records.
The inclusion of facets and other fields in the JSON response supports various usecases around automation of some research analytical tasks. As an example, the search
https://www.forskningsdatabasen.dk/en/catalog.json?f%5Bhas_orcid_b%5D%5B%5D=true&f%5Bis_active_b%5D%5B%5D=true&q=%2A%3A%2A&search_field=researchers
will query all active researchers with an ORCiD. The JSON element “total_count” is the total number of hits, and represents the (national) sum of ORCiD’s registered in any of the data suppliers local research databases. The facet “source_ss“ (JSON object “facets”, entry with “nam”equal to “source_ss”) reveals the distribution of ORCiD’s across data suppliers.