BBC News Labs Documentation

Documentation for BBC News Labs APIs & #newsHACK events

The BBC Juicer API Documentation

This is the active version of the BBC Juicer API (v2). See the BBC Juicer v1 API documentation for documentation for the old version.

About the BBC Juicer API

The BBC Juicer API allows you to search articles by text, topics, dates, and sources. It also provides aggregations and co-occurrences of people, places, organisations and 'things' as tags.

Read more about the BBC Juicer


To use this API you will need a BBC Juicer API Key. Keys are available from the BBC News Labs team.

The URL to retrieve articles{API_KEY}


Parameters can be provided in the query string.

Note: arguments must be URL encoded. For example, the url "" must be encoded as ""

Available parameters:

  • q: Keywords to search for. Searches in title, description and body of the article.

  • lang: Scopes the results to sources in the specified language. Available languages at the moment are en (English) and es (Spanish)

  • sources[]: Scopes the results to certain sources, i.e. news outlets. The parameter is a number that correspond to a source id. To specify multiple sources add multiple sources[] options, e.g. ...&sources[]=1&sources[]=5. Find the source IDs for the news outlets you're interested in via

  • facets[]: Filter the results by facets, i.e. entities from DBpedia. The parameter is a string and it refers to a DBpedia resource, e.g. "". To specify multiple facets add multiple facets[] keys and values, just like with the sources. Note: The facet parameter is URL encoded!

  • size: number of results to be returned. The parameter is a number, the default value is 25.

  • offset: number of results to skip. It can be used together with size for pagination. The parameter is a number, the default value is 0.

  • published_after: fetch articles published after a date. The date format is in UTC. Ex: 2015-02-02T00:00:00.000Z.

  • published_before: fetch articles published before a date. The date format is in UTC. Ex: 2015-02-02T00:00:00.000Z.

  • recent_first: sort results by date (with most recent first) instead of by relevance to keywords. The parameter can be either yes or empty.

  • like-text: return a list of articles with a text similar to the parameter. The like-text parameter is a string.

  • like-ids[]: return a list of articles similar to other articles based on * their textual content. The parameter is a string (the article id in the BBC Juicer database). This feature is implemented using Elasticsearch's More Like This API.

  • hist_interval: if this parameter is not empty, return a time series for the current query. The parameter specifies the granularity ot the time series and must be one of: second, minute, hour, day, week, month, year.

  • trending: if true return a list of "trending" topics for the specified date range. Trending topics are DBpedia URLs that are appearing more frequently than usual in the specified date range. Ex: published_after=2015-06-07&published_before=2015-06-08&trending=true. This uses Elasticsearch's significant terms functionality.

  • concepts: if true it will return an aggregation of all the topics found in the articles matching the current query, aggregated by Things, People, Places, Organisations.

  • sources_aggregation: if true it will return the articles count grouped by source.

All parameters are optional. If parameters are omitted, the endpoint will just return the latest articles.

That's it - Enjoy! Build great things on top of the BBC Juicer API, ask smart questions about the news! Questions and feedback to @BBC_News_Labs!

1. Get Articles by query term

Some Examples Requests

Find articles containing the word London{API_KEY}

Find articles containing London AND Festival (watch out for %20 which replaces the space in URL encoding){API_KEY}

The following query uses the search phrase Kenya OR Nairobi AND (Government OR "Uhuru Kenyatta"){API_KEY}

2. Get one Article by ID

You can get the full text of a specific article by using its id. The id is a SHA1 hash of the fully resolved URL of an article. E.g. the SHA1 (the id) for an article located at is 414adf52b72e2eac9ffd9beaf6eb02725053e9c4.


Example Request{api_key}

3. Get Similar Articles

You can get a list of similar articles by using like-ids.


Example Request[]=792c35d51f3f8d3c01bf74d940aaf4e2893fc968&api_key={api_key}

4. Get Articles from a Source

The BBC Juicer currently ingests news articles from over 400 sources (i.e. RSS feeds of news outlets). You can get a list of the available sources and corresponding IDs from the /sources endpoint.{api_key}

If you want to narrow down your search to specific sources, e.g. just the BBC (source ID is 1), you use the sources[] bit and the source ID to filter your results.

Example requests

Articles from BBC News with the query term London:[]=1&api_key={api_key}

Articles from BBC News and The Guardian with the query term Paddington:[]=1&sources[]=8&api_key={api_key}

5.Filter by topics

Topics are a way to filter down your search. They are added by our topic extraction system called Mango: it finds relevant DBpedia entities based on the text of an article. You can filter articles for the ones that have one or more of those topics by referring to its DBpedia URL.


Example Request

Get articles with the query term London that have been tagged both with David Cameron and United Kingdom.[][]{api_key}

Topics don't necessarily appear literally in the text. An article, which contains Great Britain may be tagged as United Kingdom and both Liberal Democrats and LibDems may be tagged as Liberal Democrats. The tagging process however is under development and you may discover articles talking about 'liberals' being tagged as Liberal Democrats, too. We recommend to use the topics as a discovery tool for topics that are co-occurring with certain search terms. If you want to count the articles which actually contain London and David Cameron, you could rather use a query for the two topics:{api_key}

6. Specify time interval for time-series aggregation

One common use case would be to see the appearance of news articles over time, e.g. "How frequently was the term money mentioned over the time election campaign?". Since every news article has the date attached to it you could get this information day-by-day but if you wanted to look into a longer term coverage of a topic that maybe isn't picked up so frequently (e.g. Climate Change) you can retrieve the number of articles per week or month or year, too.

Add &hist_interval= and the time unit to aggregate over to your URL. Possible time units are second, minute, hour, day, week, month, year.

Example Requests{api_key}{api_key}

7. Get a list of available Sources

Get the a list of sources currently indexed and available from the BBC Juicer. Sources are newspapers, broadcast TV channels and other sources.

Available parameters:

  • name: keyword to search for. It searches in name of the source.

Example request

Get all sources:{api_key}

Get all sources by name:{api_key}

BBC Juicer API Postman Collection

If you install the free Postman REST Client you can import the BBC Juicer API into using this URL:

This allows you to easily try out the API and explore how it works without writing any code.

You will need to configure an "Environment" within Postman to be able to make calls, add api_key as a key and your BBC Juicer API key as the value, this will replace {{api_key}} in the requests and the URL variables to your key.

Questions and Feedback

All comments are moderated for approval.