Back to search

Aristotle University Of Thessaloniki (AUTH)

Μέθοδοι στατιστικής και μηχανικής μάθησης στην ασφάλεια πληροφοριών

Abstract

dc:description

The research field of the present dissertation is the Information Security, a very special and crucial subject of the contemporary society, in which, many everyday works are conducted by using technology achievements. The study, development and application of statistical and machine learning methods, which is performed in the present dissertation, has as goal the creation of a research framework with data analysis tools that will contribute essentially to the efforts of the Information Security research community. After a first introductory Chapter, in the 2nd Chapter, the Software Vulnerabilities are presented, which constitute one of the main problems of Information Security, having intense research activity. In this Chapter, using statistical analysis, useful information is extracted regarding the vulnerability nature, that help to better understand the vulnerabilities. In the 3rd Chapter, the Vulnerability Scoring problem is described, which is a very crucial issue for the Software managers, since it defines the vulnerability prioritization. For this reason, a new vulnerability scoring system is proposed that achieves good results regarding the vulnerability accuracy and the diversity of scoring values. In the 4th Chapter, the Text Mining from the Vulnerability Description is presented. Specifically, a methodology that uses text mining techniques is proposed, in order to find correlations between terms and vulnerability characteristics/scores and to predict the vulnerability characteristics, ratings and scores. The results highlight the importance of the information that exists in the vulnerability technical description, since a) significant correlations were found from the statistical analysis and b) the estimation models are quite accurate. Finally, in the 5th Chapter, the phenomenon of the Impact of Information Security Events in the Stock Market and the prevailing methodology for studying these events, are presented. The Systematic Literature Review that is conducted on this subject, shows that generally these events are very important for the stock market, although for the recent software vulnerabilities specifically, the statistical analysis that is performed does not validate their impact to the stock market.

Degree

thesis:*
Grantor dc:publisher
Aristotle University Of Thessaloniki (AUTH)
Year dc:date
2019

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Spanos, Georgios
  • Σπανός, Γεώργιος

Subjects

dc:subject × 14

Rights

dc:rights
Statement dc:rights
  • BY_NC
Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/45101

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Spanos, Georgios; Σπανός, Γεώργιος. Μέθοδοι στατιστικής και μηχανικής μάθησης στην ασφάλεια πληροφοριών. Aristotle University Of Thessaloniki (AUTH), 2019. http://hdl.handle.net/10442/hedi/45101