Finding Relevant News Faster

was presented by IPTC at the WAN-Ifra Newsroom Summit, on Monday, 5 October 2015, 16:35 hrs at the WAN-Ifra Publishing Expo in Hamburg, Germany.

Join the presentation and we will show you:

  • how to separate relevant news from noise efficiently and effectively,
  • how to find text news, photos or graphics associated with an interesting news story
  • how software tools and editorial systems can assist you an

Background information:

The latest news continuously arrives from many different sources. Finding the content which is relevant for a publication is a key challenge for newsroom management. This process initially seems quite straight forward: define a set of characteristics for news stories and then filter the inbound feed to select the desired content. However, the day-to-day reality is that it is much harder to apply such filtering to multimedia news across text, photos, graphics or even videos arriving on a central desk. At this point the software used by a newsroom is critical to being able to achieve the desired result.

A rich and properly tagged news item delivers details about topics, persons, organisations or locations the item is about; in addition it may provide rights and administrative data such as the exact date and time of creation and if the item is embargoed. The major difference in this new generation of metadata is how the actual data is expressed. Precise identifiers can be provided for what the story is about, not simply keywords which could be miss-typed or differ based on the language, e.g. the names of persons or locations. Additionally, indications such as “this item is embargoed until 18:00” can be stated in a specific metadata field, and no longer buried in a note somewhere in the text of a news item.

These new metadata capabilities needs to be supported by the software used by a newsroom; this could be part of an editorial system or specifically in the management of incoming feeds. The software must be able to process identifiers, and add human-readable related information such as the name of a person, or the location of an event on a digital map. Additionally, such software must be able to interpret specific fields such as an embargo timestamp and provide notifications in a user-friendly way.

The essential behavior for the efficient management of incoming feeds is flexible and precise filtering. All news agencies using such a modern news format also provide predefined lists for metadata such as genre of news items and the identifiers for topics, persons and organisations; the news format then defines where these properties should be included in the news feed.

The user interface should enable the editor to easily define data about the content; this could include the geographic area of interest (e.g. a local region), typical topics which should be instantly forwarded to a specific desk (e.g. politics, economy, sports etc.) and how to deal with constraints like embargoes. Any modern editorial software should also support the latest management of updates and corrections of news items, enabling a higher level of automation to be achieved.

Ultimately the software should support the next step of journalistic work: many editors start searching for more information about a topic, person or organisation having a major role in the initial story. The software must therefore use the identifiers assigned to a story and match them against the various multimedia items provided by different sources; this enables fast retrieval of relevant background information from the content store and/or archive.


Dave Compton, Information Architect, News Platform, Thomson Reuters
Robert Schmidt-Nia, Managing Director, dpa mediatechnology
Michael Steidl, Managing Director, IPTC