E-Paper
PU conducts study to predict gender through handwriting
It is pilot study that paves way for creating larger reference databases of handwriting and using automated feature extraction; the method can be used to predict the gender of the person in tune with the requirement of the present age, where methods of forensic identification are under scrutiny for the error rates and empirical foundations
A study has been conducted by the Panjab University’s Institute of Forensic Science and Criminology that can be used to predict the gender of a writer of a document through his/her handwriting.
The study has been carried out in collaboration with Forensic Science Laboratory, Rohini New Delhi, Regional Forensic Science Laboratory (RFSL), Mandi, Himachal Pradesh, and Forensic Science Laboratory (FSL), Madhuban, Karnal, Haryana. It has been published in the Australian Journal of Forensic Sciences.
The study focuses on the identification of individual characteristics in handwriting for the purpose of estimating the gender of the writer.
A total of 150 individuals participated in the study and the relevant features were extracted from their handwriting. These characteristics were collectively studied and their potential of gender estimation was explored by utilising statistical methods of chi-square test, logistic regression, and likelihood ratios.
This is a pilot study that paves]way for creating larger reference databases of handwriting and using automated feature extraction. The method can be used to predict the gender of the person in tune with the requirement of the present age, where methods of forensic identification are under scrutiny for the error rates and empirical foundations.
The study proposes the significance of the individual characteristics of handwriting from 150 individuals from the Indian population in the estimation of the gender of the writer under question or suspicion.
Vishal Sharma of PU’s forensic science institute said, “In this study, we had different steps, starting from the extraction of individual features from the set of handwriting, converting them into data form, use of statistical method to check the significance and then application of logistic regression for model creation.”
He added, “We have checked this model with unknown samples and the correct classification rate for females and males are 80% and 76.4%, respectively, which is very high for such studies. This pilot study paves a way for creating a database of the handwriting of Indians and then automation using the current method for quick results. There is a dire necessity for such methods in countries like India where forensic identification is under scrutiny for empirical foundation and significance level in real casework.”
