查看文章

Millions of people login daily to several different internet platforms (eg, Facebook, Google, Twitter, Instagram and Reddit), which has fundamentally changed how humans communicate, seek information, discuss a plethora of topics, follow their interests, as well as a number of other activities. As a direct result of the widespread adoption of these platforms, the stream of user information being generated online has increased exponentially in the last few years. For this reason, researchers of different scientific domains have increasingly leveraged this data to investigate human behaviour. For instance, previous work has highlighted how the information contained within these social streams can provide rich insights on people’s opinions and perceptions on a number of different topics [23, 34], what motivates certain online behaviours [22, 27], and effects of social networking sites on users [20]. However, an important challenge that scientists face with such large behavioural datasets is that of information overload due to the sheer quantity of available data [19].

In this paper we describe the design and implementation of a web-based platform, called RDV (Reddit Data Visualisation), meant to provide an easy and approachable way to analyse Reddit's entire publicly available comment dataset. Reddit is an entertainment, social networking, and news website where registered community members can submit content, such as text posts or direct links,