Εντοπισμός εσωτερικών απειλών με χρήση τεχνητής νοημοσύνης σε δεδομένα κοινωνικής δικτύωσης.
Year: 2019
Type: article
Abstract: The risk of insider threat is a hidden trap that is chipping away at the security of nations and organizations who have no means to detect it.Even though technology has provided us with great strides in functionality and newer flashy systems, these systems tend to face the insider threat in a reactive rather than a proactive manner.As a direct effect of this, insider attacks are more often than not discovered after the damage has been done.Studies have shown that there are numerous ways to approach insider threat detection.One of those is affect or emotion analysis, which focuses on the emotions that are emitted during our interactions on the internet and specifically social media sites.This thesis undertakes the goal of studying this field and developing document emotion classifiers based on machine learning and social media messages exchanged on the social media site Twitter.com.Focus is provided to gathering and creating an aggregated dataset with tagged emotional tweets that will be used for training the machine learning classifiers.In addition to developing the machine learning classifiers, this this also provides a relational database schema that can be used for the creation of a complete system for insider threat detection.v Ευχαριστίες Θα ήθελα να ευχαριστήσω τον Καθηγητή μου, Δρ.Σταύρο Σιαηλή, που μου έδωσε την δυνατότητα να εργαστώ στο ενδιαφέρον αυτό θέμα, όπως και όλους όσους στάθηκαν δίπλα μου σε αυτό το ταξίδι της ζωής μου.Πάνω από όλα θα ήθελα να ευχαριστήσω την αγάπη της ζωής μου, τη γυναίκα μου, Γιώτα, που στάθηκε δίπλα μου με αγάπη, υπομονή και επιμονή και μου έδωσε δύναμη όταν εγώ δεν είχα, ώστε να φέρω εις πέρας αυτό το δύσκολο έργο.
Source: Kypseli
Author: Κυριάκου, Παναγιώτης
Language: Greek
Topic: Emotion and Mood Recognition
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6988294289
