Ever since the unveiling of the GKG 2.0 file format last month, we've been inundated with requests to expand GDELT from the news media, human rights reports, and academic literature to include global leader speeches in order to begin computationally exploring the underlying evolutionary arcs, pivots, and psychological processes expressed through the daily speeches of the world's heads of state. As a first step towards this, we are excited today to announce two new GKG 2.0 datasets that include the new GCAM pipeline, one codifying 101 speeches made by Syrian President Bashar al-Assad from 2000 to present and the second codifying a comparison dataset of 21,295 presidential documents from The American Presidency Project for Presidents Clinton, Bush, and Obama (all documents other than those with "Letter To", "Notice", "Memorandum", or "Executive Order" in the title through September 30, 2014).
Like all GKG files, the datasets below do NOT contain the text of the original documents – they contain only the codified metadata computed from those documents, including, for the first time, the output of the new Global Content Analysis Measures (GCAM) pipeline, codifying more than 2,230 emotional and thematic measures for each speech.
English translations of Syrian President Bashar al-Assad's speeches were used, drawn from "Cabayan, H. & Wright, N. (Eds) (2014).A Multi-Disciplinary, Multi-Method Approach To Leader Assessment At A Distance: The Case Of Bashar Al-Assad Parts I & II (2014). Washington, DC: Strategic Multilayer Assessment Office, Office of the Secretary of Defense." Each speech is labeled with the date it was given and "-speech1" or "-speech2" to indicate when multiple speeches were given on the same day. No other details, such as source, translator, title or venue, are available for the speeches. For further information on the speeches, please see the reference above. All speeches for American Presidents Clinton, Bush, and Obama were drawn from The American Presidency Project and include the full citation, including title, of the speech and a URL back to its entry on the project's website with further details.