Public Perception on Twitter

We use Twitter data for several projects on public perception. Most importantly, as presented in our chapter on perceptions carbon dioxide removal in the State of CDR, our goal is to develop methodologically sound methods to track and explain changes in the attention, sentiment, and emotions towards emerging climate technologies. Doing so with the help of publicly available data allows us to perform the analyses in near-real-time and on historical data. This approach complements studies in the social sciences that are based on focus groups and (large-scale) surveys and hence require longer update-intervals.

In order to process large amounts of data, we utilise two-dimensional semantic landscapes of embedded texts, state-of-the-art pre-trained classifiers, and embedding-based topic models.

So far, we have worked on, or have been involved in projects on

Datasets & Source code

  • Data and code for our work on how the onset of the pandemic temporarily shifted attention away from climate change
Tim Repke
Tim Repke