Authorship attribution and stylometric analysis of Dickens's journals

2017 Fiscal Year Final Research Report

• Project/Area Number: 15K02600
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: English linguistics
• Research Institution: Osaka University
• Principal Investigator: Tabata Tomoji 大阪大学, 言語文化研究科(言語文化専攻), 准教授 (10249873)
• Project Period (FY): 2015-04-01 – 2018-03-31
• Keywords: Dickens / style markers / authorial takeover / data mining / machine learning / stylometry / authorship attribution / mixed authorship

Outline of Final Research Achievements

Exhaustive comparative analysis has been carried out on a wide range of data mining techniques to help develop a highly accurate style variation detector on texts of mixed authorship. The development of the analytical methods draws heavily on machine-learning approaches in an effort to identify subtle stylistic shifts or variations in texts. The stylometric authorship attribution methods studied in this research have achieved a high degree of accuracy, making it possible to pinpoint where one author takes over from the other in texts of mixed authorship.

Free Research Field

コーパス文体論，デジタルヒューマニティーズ