Please use this identifier to cite or link to this item:
Title: Fake news detector
Authors: Bondin, Luca
Keywords: Machine learning
Fake news
Graph theory
Issue Date: 2018
Citation: Bondin, L. (2018). Fake news detector (Master's dissertation).
Abstract: With the amount of information available on the web and the ease with which everyone can add his/her own input, the need for tools that help in highlighting only the truthful sections of what is being made available has never been greater. The proposed system, is envisaged to help the common user distinguish between what can be classified as real or fake news through the use of an easily accessible online tool. The application focuses on using the best techniques to help identify truthful articles from others more commonly referred by the term “fake news”. The project considers the current state of the art and aims to going a step further in integrating various technologies in order to achieve the best possible results by proposing a tripartite approach (using Machine Learning, Graph Analysis and NLP techniques). Each separate component is evaluated first in isolation and then in unison with the other components in order to obtain a more in depth understanding of the system’s performance. Tests have shown that the overall accuracy of the system in determining the nature of a news article lies at 85%. Future work on this study opens the door to the implementation of similar systems that are to be deployed in the field of deception detection at a time where the need of such systems is all the more increasing.
Appears in Collections:Dissertations - FacICT - 2018
Dissertations - FacICTAI - 2018

Files in This Item:
File Description SizeFormat 
  Restricted Access
1.58 MBAdobe PDFView/Open Request a copy

Items in OAR@UM are protected by copyright, with all rights reserved, unless otherwise indicated.