Automated journalism is the process by which computers select the stories consumers see, and the order in which they see them. Nearby location can make news stories seem more relevant to consumers (following Galtung and Ruge, 1965), but a more nuanced view of locational relevance than just a street address or zip code is required to drive the most compelling stories to consumers (such basic location information is deemed "lackluster" by Weiss (2014, p. 4)). What factors could be used to enrich a computer's definition of a consumer's location to deliver the most interesting stories him or her?
If you think about reading news on smart phones, as opposed to computers, most people have "location services" turned on, so that an app can read the exact geographic coordinates, within a few feet.
Unless the automated information is very specifically designed for me and my interests, I usually don't find it that helpful. For example, in Manoa, zip code 96822, all of the campus information would be filtered to me, which is one thing, but also all of the community information as well, which is another. I might be interested in both, but, then again, I might not. So, depending on the distinctions, this all might be very useful information of a lot of noise in the system. A few years ago, I was working at a newspaper that decided to geolocate all of its stories on a county map. The system automatically did so, but, because the newspaper based most of its stories on city and county government, which happened to reside in the same few blocks of the city, the system geotagged hundreds and hundreds of stories all in the same place. So when you looked at the map, it appeared that the newspaper only covered a two-block area of the county (which, unfortunately, wasn't that far from the truth). But, the reality of the situation was that even those stories had ties to the county courthouse or city hall, they really affected a lot of other people all around the community, so these stories could have been geotagged in other ways, with a better meta-tagging system, or better reporting (with more diverse sources), or any other number of ways. What was funny about this situation (not really funny for the newspaper or the digital map developer, though) was the appearance this automated system portrayed, based in a somewhat skewed reality of dumb intelligence. The machine was smart enough to figure out that the county courthouse was located on a certain block in the city, and, therefore, it could tag the story to that spot, but it was not smart enough to see the broader connections the court case, or council vote, or whatever, had to the rest of the community.
Consumers could have stories catered to their interests and subjects that they enjoy reading. I would love to have stories that relate to my research interest in mental illnesses, social media networks, dogs, and so on. I've noticed that Facebook does feature stories related to my interests, but would appreciate if other platforms could do similar functions as well.
- Badges
- Science topic
- Similar topics
- Social Science
- Media
- Journalism