Search results

The purpose of this study is to identify the most frequent themes related with social media governance in government by year, analyzing if these themes have evolved over time, as well as highlighting the main risks and challenges found as further research opportunities.

First, the authors have extracted 431 abstracts from Scopus database. Then, abstracts were grouped by year to apply topic modeling to discover the underlying topics. Specifically, the authors have applied latent Dirichlet allocation algorithm to identify the most frequent topics by year.

The results reveal 19 important topics related with social media governance in government. Then, these topics were assigned to each year to identify the evolution of the research themes over the years, proposing interesting avenues for further research based on the identification of the main risks and challenges.

The proposed research methodology can be applied not only for research purposes but also to discover themes in any discourse with applications in politics, marketing, business, etc. In addition, it can be used to save time and costs analyzing citizen comments in public debates to identify the most important topics.

This study can serve to highlight gaps in the literature, opening the possibility that researchers can adequately position their inquiries, as well as to be aware of overstudied themes to pay less attention to them in future projects. In addition, the results of this study could serve as a starting point for other researchers to analyze connections between topics, propose theories that explain what was found and validate them in future studies.