As a public agency responsible for getting people into work, NAV keeps a job ads database of open positions in Norway. As a public actor, NAV is also required to publish available data to other actors.
Publishing data about available jobs for the Norwegian population and employers is found in this API
Terms of service for the use of the feed API can be found here: https://arbeidsplassen.nav.no/vilkar-api-gammel
The feed API requires authentication using the Authorization header, type
Bearer and a signed JWT token.
There are two types of tokens, public and private. The public token may be revoked at any time and replaced by another one. If the consumer needs a more stable token one can be ordered by mailing a request for a token to plattform.for.arbeidsmarkedet@nav.no. Please include your name or the company name in the request.
The current public token is
eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJzdWIiOiJwdWJsaWMudG9rZW4udjFAbmF2Lm5vIiwiYXVkIjoiZmVlZC1hcGktdjEiLCJpc3MiOiJuYXYubm8iLCJpYXQiOjE1NTc0NzM0MjJ9.jNGlLUF9HxoHo5JrQNMkweLj_91bgk97ZebLdfx3_UQ
Requests and responses are documented with swagger. When testing with swagger feel free to use the token above.
It is important to respect ad expiry if your store the ads in your own solutions. If an ad is past its 'expires' date/time, then it should no longer be published or visible. (NAV keeps ads published up to, but also including, the last day of expiry.)
Ads can also be "stopped" or unpublished before the set expiry time, for various reasons. This condition can be detected simply by assuming that any ad which is no longer available from the feed should no longer be published. But in order to do this, you must apply a method that ensures you have all currently active ads, which is described in the next section.
There is a limit on the number of ads available in the feed, for any particular set of filtering criteria. This limit applies to the total number of hits regardless of paging, and it is currently at 5000 ads. When no filtering criteria are used, the API will provide the 5000 most recently published ads, but no more. The feed is always sorted on publishing date, from most recent to older.
If you need to get all currently active ads, there are a few options:
-
Filter on published-date. If you start by consuming the entire feed the first time, you can "page back in time" by filtering subsequent feed page-throughs on the oldest published-date of the previous fetch (e.g. the published date of the last ad). This way, you get the next 5000, which will all be older than the ads in the first batch. Continue going back in time until you get no more hits. Then you will have fetched all currently active ads.
As an example, if the oldest ad of the first fetch (without any filtering criteria) has a published date of 2019-06-01T00:30:00, then for the next fetch, the following filter criteria could be used:
/public-feed/api/v1/ads?published=[*,2019-06-01T00:30:00)And so on.
Also, it is important to realize that this method assumes that you page through all available ads until none are left, everytime the published filter criterium is adjusted.
-
Consume the feed regularly (with full page-through) and store all ads in your own solutions. You will get all the new ads since the last time the feed was fetched. You can always rely on the UUID field to determine if an ad is new or you already have it. (It can also be a good idea to update data on ads that you have already stored.) There are approximately 1000 new jobs ads published per day, but it can vary quite a bit. After a few days of doing this, you will have a complete set of all active ads. (Also remember to unpublish expired ads that you have stored, see previous section.)
-
Future option: the feed will eventually allow filtering on geography and possibly other data, which can be used to bring the complete number of matching ads well below the limit of 5000. In that case, you can assume you have all relevant ads after one feed page-through.