CODE BREAKING APPROACH IN AUTOMATED CHARACTER RECOGNITION
Journal Title:Globus : An International Journal of Management and IT
The Code Breaking Approach in Automated Character Recognition is a system that tests the security of free social networking web site CAPTCHA images. It compares the efficiency and accuracy of two methods, the parts-based and segmentation which serve as a guide and also a basis for further studies in AI community. The researcher has employed both experimental and descriptive design by using online generated data samples which is used for gathering accuracy and efficiency result of parts-based and segmentation method. The findings shows that the over-all result in comparing the accuracy and efficiency of two methods have no significant difference using the file size range of a CAPTCHA image. On the other hand, the study still got efficient and accurate results which show that the system is reliable in testing the accuracy and efficiency in identifying image code using the two methods. Therefore, the actual implementation of the Code Breaking Approach in Automated Character Recognition shall be recommended to be used in improving internet security, especially in free email services and social networking websites that provides security and authentication to its user.