International Journal of Computer Applications
Foundation of Computer Science (FCS), NY, USA
|
Volume 184 - Issue 1 |
Published: Mar 2022 |
Authors: Dian Pratiwi, Syaifudin, Muhammad Azamy |
![]() |
Dian Pratiwi, Syaifudin, Muhammad Azamy . Modified Chi-Square Distance to Improve Personality Type Recognition based on Handwriting. International Journal of Computer Applications. 184, 1 (Mar 2022), 6-12. DOI=10.5120/ijca2022921915
@article{ 10.5120/ijca2022921915, author = { Dian Pratiwi,Syaifudin,Muhammad Azamy }, title = { Modified Chi-Square Distance to Improve Personality Type Recognition based on Handwriting }, journal = { International Journal of Computer Applications }, year = { 2022 }, volume = { 184 }, number = { 1 }, pages = { 6-12 }, doi = { 10.5120/ijca2022921915 }, publisher = { Foundation of Computer Science (FCS), NY, USA } }
%0 Journal Article %D 2022 %A Dian Pratiwi %A Syaifudin %A Muhammad Azamy %T Modified Chi-Square Distance to Improve Personality Type Recognition based on Handwriting%T %J International Journal of Computer Applications %V 184 %N 1 %P 6-12 %R 10.5120/ijca2022921915 %I Foundation of Computer Science (FCS), NY, USA
This research was conducted to develop a mobile device that is able to recognize one’s personality to support expert decisions based on handwriting through the application of graphology and enneagram psychology. In the process, handwritten data is processed in three main stages, namely pre-processing, texture feature extraction in the form of contrast, energy, and entropy with GLCM, and similarity measure through the modified chi-square method. The value of the feature is categorized into 4 categories in the form of slant, size, breaks and baseline, which will be stored in the SQLite database as a reference. Later, the determination of personality will be seen based on the calculation of the smallest distance of the test data on the reference value of the combination of the categories. Based on the results of the study, the three GLCM texture feature values obtained have intervals that are not unique, and difficult to distinguish between one personality type to another. But the use of the modified chi-square method in the form of random weights can process the feature values so that the test data can be distinguished by type personality with each other with a precision of 60-80% and an accuracy of 72%.