Scratch as Social Network: Topic Modeling and Sentiment Analysis in Scratch Projects
share
Societal matters like the Black Lives Matter (BLM) movement influence software engineering, as the recent debate on replacing certain discriminatory terms such as whitelist/blacklist has shown. Identifying relevant and trending societal matters is important, and often done using social network analysis for traditional social media channels such as twitter. In this paper we explore whether this type of analysis can also be used for introspection of the software world, by looking at the thriving scene of young Scratch programmers. The educational programming language Scratch is not only used for teaching programming concepts, but offers a platform for young programmers to express and share their creativity on any topics of relevance. By analyzing titles and project comments in a dataset of 106.032 Scratch projects, we explore which topics are common in the Scratch community, whether socially relevant events are reflected and how how the sentiment in the comments is. It turns out that the diversity of topics within the Scratch projects make the analysis process challenging. Our results nevertheless show that topics from pop and net culture in particular are present, and even recent societal events such as the Covid-19 pandemic or BLM are to some extent reflected in Scratch. The tone in the comments is mostly positive with catchy youth language. Hence, despite the challenges, Scratch projects can be studied in the same way as social networks, which opens up new possibilities to improve our understanding of the behavior and motivation of novice programmers.
READ FULL TEXT