Photography at the Central Criminal Court

This dataset contains the results of a project comparing the results of human and AI-generated thematic analyses of trials relating to photography from the Old Bailey Proceedings. It comprises six fields: ‘Year’, containing the year of the trial; ‘Trial ID’, containing the ID assigned by the Old Bailey Proceedings Online (OBO); ‘Offence’, containing the OBO code for the relevant offence; ‘Manual Categories’, containing a semicolon-separated list of categories generated and assigned by a human researcher; ‘Auto Categories from Manual List’, containing a list of human-generated categories assigned to the trial by GPT-4o (20/11/2024 version); and ‘AI-Generated Categories’, containing a list of categories generated and assigned by GPT-4o. Code and prompts used to interface with GPT-4o can be found at the associated Github link.