**Intra-textual sentiment analyses of lone actor manifestos**
This project analyses the sentiment of sentences throughout a dataset of violent lone actor manifestos. Our aim is to discover how and what these individuals write regarding the violent attack they committed. Specifically, we examine fluctuations in positive and negative language use and the concepts that give rise to these sentiments. The dataset consists of 22 manifestos across a range of attack motivations (far-right, jihadist, pro-life).
Linguistic features datasets are available on this page, as well as the code for analysis. Raw text files can be obtained upon request by e-mailing us.