Writing a one column news story using the five Ws and one H — Who, What, Where, When, Why and How — is pretty routine for most journalists. These news stories are not long and are written using a fixed template. You could use this template again; just replace the data for a new story. Can you teach this trick to a computer? Writing by the computer is a matter of using a good algorithm; using certain words in sentences in the paragraph order of the news article. However you have add the data yourself, unless these data come also from a computer.
On Monday 17 March, the LA Times reported that an earthquake with a magnitude of 4.4 on the Richter scale shook California. The story was written by Quakebot, developed by Ken Schencke, programmer at the newspaper, reports Will Oremus of Slate. Earthquakes are recorded by the US Geographical Survey(USGS). When a quake occurs, the data is picked up by the Quakebot, which uses this information in a predefined template to write the story and stores it in the newspaper’s content management system. Here is the report in full:
A shallow magnitude 4.7 earthquake was reported Monday morning five miles from Westwood, California, according to the U.S. Geological Survey. The temblor occurred at 6:25 a.m. Pacific time at a depth of 5.0 miles.
According to the USGS, the epicenter was six miles from Beverly Hills, California, seven miles from Universal City, California, seven miles from Santa Monica, California and 348 miles from Sacramento, California. In the past ten days, there have been no earthquakes magnitude 3.0 and greater centered nearby.
This information comes from the USGS Earthquake Notification Service and this post was created by an algorithm written by the author.
At the same time, the algorithm sends an e-message to the human editor, who can publish the story with push of a button. Of course, not all earthquakes are handled in this way; a magnitude less than three is too small to have an impact and us therefore neglected.
The Quakebot is not however completely replacing journalists. The first story written by the algorithm is just the opening of a developing news story. Journalists can then start their investigations, checking damage or interviewing sources about their experiences.
The original quake story was updated 71 times and by the end of the day had become an extended news story based on human reporting. The Quakebot is therefore an interesting supplement to the work of journalists. It works as a news alert which can be published immediately and followed up in due course.
This isn’t the first time the LA times has experimented with automated journalism. One of the most well-known ones is the homicide report, a crime map of the latest homicide in LA based on police data, showing where and how murders have taken place, regularly accompanied with a picture of the victim.
This is not a news article but a map with pinpoints and data which tell the story about the latest homicide. These pinpoints on the map could however become the beginning of a developing news story. The homicide report started seven years ago is and since 2009 has helped drastically change the balance of crime coverage at the newspaper.
It’s not just information around earthquakes and murders that immediately be transformed into a standardised news story though; think about unemployment figures, sports results or financial data. Forbes for example employs algorithms developed by Narrative Science to write automatic financial news stories. Automated reporting is possible as long as the data fits in the set of rules of the algorithm. Robots in the newsroom are faster and save time for the human reporter to do the actual journalistic work.
There are though some challenging questions that need to be answered. For example, who is liable in case the bot makes a mistake? A journalist always checks his data, but the bot does not. It could, for example, report that a policeman was shot dead, only for it to later turn out that he was severely wounded. The law will follow later.
Robots in the newsroom are the beginning of a new development in journalism. Data journalism is already changing what we mean by in-depth reporting. Thing is, the amount of data is growing and journalists need a first report, a news alert, from this data in order to start an investigation. The growing mountain of data can only be an incentive for automated journalism and an increased number of robo journos
Bots and other cutting-edge tools used by innovative journalists today will be the topic of a discussion panel at the 15th International Symposium on Online Journalism (ISOJ), titled “Bots, drones, sensors, wearables, etc.: The new tools for journalists.” ISOJ will take place on April 4 and 5 at the University of Texas at Austin.
Image: Lindsey Turner (via Flickr)