The final panel on this day at AoIR 2018 is on journalism, and starts with Òscar Coromina. His focus is on the influence that trending topics on Twitter had on journalistic coverage of the Catalan independence referendum. Trending topics are important in directing user attention, especially in the context of breaking news, and Twitter is of course also selling advertising at the top of its trending topics list, indicating their importance.
Trending topics may be hashtags or phrases, and work in similar ways to enable the formation of ad hoc publics or algorithmically generated publics; they are technosocial actors in the mediation of different events, and indeed one of the spaces on Twitter where algorithmic curation is most obvious.
How do such trending topics appear and disappear, then? The project used DMI-TCAT to capture tweets matching a manually updated list of keywords relating to the Catalan independence referendum. These included relevant hashtags, categorised as facilitating description, emphasis, calls to action, and media coverage. These were mainly local (Barcelona), national (Spain), or both; few were international. Media hashtags were most prominent, followed by emphasis; the other categories were less prominent.
While trending topics largely represented media content, the largest number of tweets were attracted by descriptive hashtags. Hashtags largely fell into disuse within a few hours (call to action hashtags persisted for a few days). Trending topics lists are therefore a space of competition and dispute in which different actors fight for attention and framing in a highly accelerated news cycle. They are difficult to capture using tracking approaches that use a static list of keywords. Trending topics were largely connected with media coverage, and this indicates the increasing attention that media coverage now pays to such topics.
One of the key issues here is the tracking of trending topics; discovering which topics are currently trending is not trivial, and there is no direct mechanism for feeding trending topics lists to Twitter trackers at this stage.