Spotting political social bots in Twitter: a dataset for the 2019 Spanish general election
Description
While social media has been proved as an exceptionally useful tool to interact with other people and massively and quickly spread helpful information, its great potential has been ill-intentionally leveraged as well to distort political elections and manipulate constituents. In the paper at hand, we analyzed the presence and behavior of social bots on Twitter in the context of the November 2019 Spanish general election. Throughout our study, we classified involved users as social bots or humans, and examined their interactions from a quantitative (i.e., amount of traffic generated and existing relations) and qualitative (i.e., user's political affinity and sentiment towards the most important parties) perspectives. Results demonstrated that a non-negligible amount of those bots actively participated in the election, supporting each of the five principal political parties. The dataset at hand presents the data collected during the observation period (from October 4th, 2019 to November 11th, 2019). It includes both the anonymized tweets and the users' data. Instructions ------------- Data have been exported in three formats to provide the maximum flexibility - MongoDB Dump BSONs: To import these data, please refer to the official MongoDB documentation. - JSON Exports: Both the users and the tweets collections have been exported as canonical JSON files. - CSV Exports (only tweets): The tweet collection has been exported as plain CSV file with comma separators.
Files
Steps to reproduce
Code is fully available on the official Github repository, linked in this page.