Back to search

University of the Aegean

Authorship verification

Abstract

dc:description

Authorship analysis attempts to reveal information about authors of digital documents enabling applications in digital humanities, text forensics, and cyber-security. Author verification is a fundamental task in authorship analysis. Given a set of documents, all by the same author, the question is whether another disputed document is also by the same author. Authorship verification has gained a lot of attention during the last years mainly due to the focus of PAN@CLEF shared tasks. In this dissertation, we propose improved author verification methods and systematically examine their performance on several benchmark datasets.First, we propose two enhanced algorithms, one instance-based (all documents of known authorship are treated separately) and one profile-based (all known documents are treated cumulatively) that are able to outperform state-of-the-art methods in several benchmark datasets. We demonstrate that the proposed methods are capable of taking advantage from the availability of multiple documents of known authorship and they are robust when text length is limited.Moreover, we explore the usefulness of topic modeling in author verification. We conduct a systematic study to examine whether topic modeling techniques achieve to improve the performance of certain author verification paradigms and what particular topic modeling technique is the most appropriate for each individual author verification paradigm. To this end we combine the well-known Latent Semantic Indexing or Latent Dirichlet Allocation topic modeling methods with several author verification methods that cover the main paradigms in this area, namely intrinsic (attempt to solve a one-class classification task) and extrinsic (attempt to solve a binary classification task) methods as well as profile-based and instance-based approaches. We use benchmark datasets and demonstrate that LDA is better combined with extrinsic methods while the most effective intrinsic method is based on LSI. Moreover, topic modeling seems to be particularly effective for profile-based approaches and the performance is enhanced when latent topics are extracted by an enriched set of documents. The comparison to state-of-the-art methods demonstrates the great potential of the approaches presented in this study. It is also demonstrated that even in case genre-agnostic external documents are used, the proposed extrinsic models are very competitive.In some of the relevant studies, there is evidence that heterogeneous ensembles can provide very reliable solutions, better than any individual author verification model. However, there is no systematic study of examining the application of ensemble methods in this task. We attempt to fill this gap by starting from a large set of base verification models covering the main paradigms in this area and study how they can be combined to build an accurate ensemble. We propose a simple stacking ensemble as well as a dynamic ensemble selection approach that can use the most reliable base models for each verification case separately. The experimental results in ten benchmark corpora covering multiple languages and genres verify the suitability of ensembles for this task and demonstrate the effectiveness of our method, in some cases improving the best reported results by more than 10%.

Degree

thesis:*
Grantor dc:publisher
University of the Aegean
Year dc:date
2019

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Potha, Nektaria
  • Πόθα, Νεκταρία

Subjects

dc:subject × 13

Rights

dc:rights
Statement dc:rights
  • BY_NC_ND
Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/45382

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

Potha, Nektaria; Πόθα, Νεκταρία. Authorship verification. University of the Aegean, 2019. http://hdl.handle.net/10442/hedi/45382