Please use this identifier to cite or link to this item: http://hdl.handle.net/10609/124346
Title: Detecció de bots en xarxes socials per mètodes supervisats
Author: Consuegra Navarrina, Josep
Director: Solé-Ribalta, Albert  
Tutor: Vicens Bennasar, Julian Antonio
Abstract: This project is born as a result of the public opinion vulnerability in regard to social networks, where bot presence, main responsible of fake news propagation and misinformation spread, act with impunity by taking advantage of non-existing or inefficient bot detection (and control) protocols. The goal of this project is, mainly, to implement a binary classification algorithm for Twitter users, in charge of detecting whether a user is behaving as a bot or not. The algorithm is based on a user activity dataset consisting of 650k tweets downloaded through the Twitter API between April 24th and May 5th, as well as a training dataset obtained by using Botometer API. Only supervised methods are considered for the implementation, based on the users' activity in Twitter (without considering the contents of the tweet's body), which are afterwards compared, showing that MLP and Random Forest classifiers seem to perform better in this scenario. For visualization purposes, all users from the original dataset are then classified as a human or a bot, and are added into a graph, where each node represents a user and each edge represents an interaction. Additionally, a community detection algorithm is applied, and the graph is visualized through Gephi tool, showing that there is a polarization of users, and that bots seem to be equally distributed among all communities, meaning they are inherent to the network.
Keywords: social networks
misinformation
graphs
Document type: info:eu-repo/semantics/masterThesis
Issue Date: Jun-2020
Publication license: http://creativecommons.org/licenses/by-nc-nd/3.0/es/  
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
Consuegra_tfm.pdf1,2 MBAdobe PDFThumbnail
View/Open
jconsuegranTFM0620memòria.pdfMemòria del TFM1,15 MBAdobe PDFThumbnail
View/Open