ID | 114296 |
Title Alternative | MACHINE LEARNING APPROACH FOR AVOIDING RAPE
|
Author |
Yokotani, Kenji
Niigata Prison|Niigata Seiryo University
Tokushima University Educator and Researcher Directory
|
Keywords | Criminal suit documents
Supervised machine learning
Binary classification
Protective action
Rape
Sexual coercion
|
Content Type |
Journal Article
|
Description | Effective self-protective behaviors, such as victim's physical resistance for avoiding sexual victimization have been studied. However, effective self-protective behavioral sequences, such as offender's physical violence followed by victim's physical resistance, have not been studied often. Our study aims to clarify these sequences through supervised machine learning approach. The samples consisted of 88 official documents on sexual crimes regarding women committed by male offenders incarcerated in a Japanese local prison. The crimes were classified as completed or attempted cases based on judges’ evaluation. All phrases in each crime description were also partitioned and coded according to the Japanese Penal Code. The Support Vector Machine learned the most likely sequences of behaviors to predict completed and attempted cases. Around 90% of cases were correctly predicted through the identification of sequences of behaviors. The sequence involving the offender’s violence followed by victim’s physical resistance predicted attempted sexual crime. However, the sequence involving victim’s general resistance followed by the offender’s violence predicted completed sexual crime. Timing of victim’s resistance and offender’s violence could affect potential avoidance of sexual victimization.
|
Journal Title |
Asian Journal of Criminology
|
ISSN | 18710131
1871014X
|
NCID | AA12248315
|
Publisher | Springer Nature
|
Volume | 13
|
Issue | 4
|
Start Page | 329
|
End Page | 346
|
Published Date | 2018-07-27
|
Remark | This is a post-peer-review, pre-copyedit version of an article published in Asian Journal of Criminology. The final authenticated version is available online at: https://doi.org/10.1007/s11417-018-9273-1.
|
EDB ID | |
DOI (Published Version) | |
URL ( Publisher's Version ) | |
FullText File | |
language |
eng
|
TextVersion |
Author
|
departments |
Integrated Arts and Sciences
|