It is time to break-down the broad concept of “data journalism”. When talking about the combination of data and news, we usually refer to two processes, sometimes conducted in an integral manner. One process is to discover news points from datasets. The datasets can provide a lead for further investigation. The final product does not necessarily reflect the usage of data. It may look the same as normal news products mainly composed of interviews and photos. This is called “data mining” in the science domain. Another process is to present news points using data. There come to all kinds of charts and interactive/ immersive presentations. This is called “data visualisation” in the science domain.

Let’s focus on the “data mining” part in this article. That is to discover news from datasets, or more precisely discover a news lead from datasets. The further development of the entire news story may take much more efforts with a combination of traditional and modern methods. For easier discussion, we treat “news” in the general form: something the audience does not know before reading, a.k.a, something that “appears new”. It could be the status update of a current affair, or it could be the “new knowledge” to the readers (probably be “common knowledge” to experts which we don’t want to waste time debating).

As advocated by the “Road to Jan”: the most profound theory takes the simplest form. As a first step, we try not using programming, or even sophisticated spreadsheet skills. One can readily find some “news” with a bit “nose for news” and be computer literate is good enough. In this article, we will demo a few news points mined by our undergraduate students from Hong Kong government data portal: https://data.gov.hk . It took around 20 minutes in the second class of a data journalism course. We start with a public dataset from the portal, check out the data tables and eyeball if there is anything interesting. The process is so quick that we would like to give it a brand name: Lightning News. One can sharpen his/her news sense and data sense by doing this as daily exercise.

Continue reading →