Mining and analysing online social networks: Studying the dynamics of digital peer support

In recent years, the rapid growth of user-generated content has led to much research evaluating the patterns of online information exchange. These studies demonstrate that online communities are valuable data sources which provide rich, longitudinal data that would otherwise be difficult, if not impossible to access. Given the increased research interest, mining and analysing online social networks has become an important research domain, encompassing a variety of approaches. To analyse the large number of observations commonly found in online communities, we propose to first mine the data using a so-called Webscraper and then combine Social Network Analysis (SNA) with Sentiment Analysis to explore both content and relationships. The hands-on approach described in this article is targeted at researchers without a background in technical disciplines. Instead of focusing on some of the specific algorithms that facilitate the mining and analysis of online data, we describe how to use and combine out-of-the-box solutions to collect and analyse the online network data. Moreover, we document the steps taken and present important lessons learnt throughout the process of collecting and analysing data from an online health community with 108,569 registered users who contributed to 197,980 discussions with a total of 484,250 replies. In sum, our method proposes to: • Extract all relevant data from an openly accessible online community using a Webscraper. • Determine and visualise the relationships between users and the properties of the social network as a whole using Social Network Analysis. • Conduct Sentiment Analysis to detect the emotional tone of the online contributions, and to possibly infer further variables from the text such as the personality characteristics of users.