Interactive Visual Analytics‌ Sample Clauses

Interactive Visual Analytics‌. The main challenge in browsing and visualisation of interlinked media and social network content is in providing a suitably aggregated, high-level overview. Timestamp-based list interfaces that show the entire, continuously updating stream (e.g. the Twitter timeline-based web interface) are often impractical, especially for analysing high-volume, bursty events. For instance, during the royal wedding in 2011, tweets exceeded 1 million. Similarly, monitoring long running events across media, places, and time is equally complex. One of the simplest and most widely used visualisations is word clouds. These generally use single word terms, which can be somewhat difficult to interpret out of context. Word clouds have been used to assist users in browsing social media streams, e.g blog content (Bansal, 2007) and tweets (Xxxxxxxxxxx et al, 2009; Shamma, 2010). The main drawback of cloud-based visualisations is their static nature. Therefore, they are often combined with timelines showing keyword/topic frequencies over time (x.x. Xxxxxxx-Xxxxxxxxx, 2012; Xxxxx, 2011), as well as methods for discovery of unusual popularity bursts (Bansal, 2007). In addition, some visualisations try to capture the semantic relatedness between topics in the media. For instance, BlogScope (Bansal, 2007) calculates keyword correlations, by approximating mutual information for a pair of keywords using a random sample of documents. Another example is the information landscape visualisation, which conveys topic similarity through spatial proximity (Xxxxxxx-Xxxxxxxxx, 2012). Lastly, some visualisations have tried to convey the social context. For instance, the PeopleSpiral visualisation (Dork, 2010) plots Twitter users who have contributed to a topic (e.g. posted using a given hashtag) on a spiral, starting with the most active and ‘original’ users first. User originality is measured as the ratio between the number of tweets authored by the user versus re-tweets made.