Collection
Revisiting Authorship Attribution
Stylometric authorship attribution research aims to reveal the connection of a text of unknown authorship to a specific author using a set of quantifiable text features as indicators of the author’s style. It is one of the oldest applications of quantitative methods in linguistic data with relevant studies based on manual counting of linguistic features dating back to the 15th century.
Since the late 1990s, authorship attribution has known a new impetus based on developments in several key research areas such as Information Retrieval, Machine Learning, and Natural Language Processing. Furthermore, the machine-readable text is now massively available. Web 2.0 has added to the now standard internet genres of email, web page, and online forum messages, new forms of online expression such as blogs, tweets, and instant messaging. Moreover, since 2014, the NLP community was revolutionized by the rise of the word embeddings and the deep neural document representation models like BERT, which offered holistic textual representation models for all text mining tasks.
Authorship analysis research is now concerned not only with problems in the broad field of the Humanities (Literature, History, Theology) but also with applications in various law-enforcement tasks such as Intelligence, Forensics, etc. At the same time, a number of different research questions have been raised, including issues of author profile (gender, age, personality, etc.) that formed a broader research agenda and fostered the development of the broader field of computational stylistics.
In this rapidly changing research environment, authorship attribution methodology follows the evolution of text mining methods closely. The aim of this special issue is to capture the state-of-the-art in the field and give a broad coverage of the most active research areas.
We invite proposals from scholars working on authorship attribution research regardless of application domain or language(s) involved. We are interested in approaches that address challenging aspects of authorship attribution and offer novel ideas to the field advancing the current state of the art. Topics can be related but are not limited to:
Authorship attribution in the wild. Models and methods for big datasets
Author verification
New machine learning algorithms suited for authorship analysis research
Software for authorship analysis research
Authorship analysis in forensics
Authorship attribution in small datasets
Identification of novel features
Cross-linguistic and cross-topic authorship analysis
This collection was curated by the Editor in Chief from articles that also appear in the journal's issues. The journal's standard peer review policy applies here. If an article was also included in a special issue of the journal, please see the instruction for authors for the special issue peer review policy.
Editors
-
George Mikros
National and Kapodistrian University of Athens, Greece George K. Mikros is currently Professor of Computational and Quantitative Linguistics and Chair of the Department of Italian Language and Literature in the National and Kapodistrian University of Athens and Director of the Computational Stylistics Lab. He also holds an adjunct faculty position at the Department of Applied Linguistics, in the Master of Arts Online Program of the University of Massachusetts, Boston.
-
Patrick Juola
Patrick Juola currently works at Duquesne University in Pittsburgh, USA.
-
Maciej Eder
Maciej Eder is an associate professor at the Pedagogical University of Kraków, Poland, and at the Institute of Polish Language of the Polish Academy of Sciences. He is a literary scholar interested in early modern Latin and Polish literature, and a stylometrist involved in a number of projects. His work is now focused on a thorough re-examination of current stylometric methods and applying them to non-English languages, e.g. Latin and Ancient Greek. He introduced a couple of new methods, e.g. the Rolling Stylometry technique, which is aimed at analyzing texts sequentially.
