МЕТОДИ ГЕНЕРАЦІЇ ТА АНАЛІЗУ СТІЙКОСТІ ПА...
Saved in:
| Title: | МЕТОДИ ГЕНЕРАЦІЇ ТА АНАЛІЗУ СТІЙКОСТІ ПА... |
|---|---|
| Alternate Title: | METHODS OF GENERATION AND ANALYSIS OF PASSWORD STRENGTH TAKING INTO ACCOUNT LEXICAL, STRUCTURAL AND HEURISTIC CHARACTERISTICS. |
| Authors: | Кришталь, Т. Ю.1, Троянський, О. В.1 o.v.troyanskiy@op.edu.ua, Кушніренко, Н. І.1 kushnirenko@op.edu.ua |
| Source: | Informatics & Mathematical Methods in Simulation / Informatika ta Matematičnì Metodi v Modelûvannì. 2025, Vol. 15 Issue 2, p218-228. 11p. |
| Subjects: | Password software, Information technology security, Heuristic, Machine learning, Internet security, Electronic authentication |
| Abstract: | The article discusses the developed methods for generating and analyzing password strength, taking into account lexical, structural, and heuristic features, which were implemented within the framework of the development of the KrystalLock software product. The main goal of the work was to create effective methods that ensure both generation and verification of passwords for compliance with modern security requirements. Within the framework of the implemented approach, three methods of password generation are proposed: random, mnemonic, and combined. Mnemonic generation is based on creating passwords from phrases that are easy for the user to remember, with the subsequent use of selective leet transformations to increase complexity. All methods support parameter settings, in particular, password length, the use of separate categories of characters (upper/lower case, numbers, special characters), as well as control over meaningful templates. In the part of password strength analysis, a multi-level method has been implemented that combines classical complexity assessment criteria with modern approaches to machine analysis. In particular, the length, character composition, pattern detection (repetition, sequences, common keyboard combinations), checking for the presence of known merged password databases, as well as lexical and semantic analysis are carried out. Special attention is paid to the machine approach to weak password detection, which is built on the bidirectional long-term memory model (BiLSTM). The model is trained on a dataset of passwords containing various names and words and demonstrates accuracy of over 92% accuracy on test samples, using the F1, Precision and Recall metrics. The developed application is implemented in the Python environment using the TensorFlow, NumPy, secrets and other libraries. Comprehensive testing of the developed methods was carried out, including both functional and load scenarios. The compliance of the analysis results with the expected cybersecurity standards (NIST, OWASP, ISO/IEC 27001) was assessed and the effectiveness of the approach was confirmed compared to existing analogues. The results of the work can be used to increase the level of account protection in personal and corporate environments, as well as for implementation in password management systems. The application can become the basis for further research and improvement in the field of adaptive authentication, in particular by integrating behavioral analysis or biometric factors. The complexity of the approach, the use of deep learning models and the possibility of personalization make the developed product a valuable contribution to the field of practical cybersecurity. [ABSTRACT FROM AUTHOR] |
| Copyright of Informatics & Mathematical Methods in Simulation / Informatika ta Matematičnì Metodi v Modelûvannì is the property of Odessa Polytechnic University and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
| Abstract: | The article discusses the developed methods for generating and analyzing password strength, taking into account lexical, structural, and heuristic features, which were implemented within the framework of the development of the KrystalLock software product. The main goal of the work was to create effective methods that ensure both generation and verification of passwords for compliance with modern security requirements. Within the framework of the implemented approach, three methods of password generation are proposed: random, mnemonic, and combined. Mnemonic generation is based on creating passwords from phrases that are easy for the user to remember, with the subsequent use of selective leet transformations to increase complexity. All methods support parameter settings, in particular, password length, the use of separate categories of characters (upper/lower case, numbers, special characters), as well as control over meaningful templates. In the part of password strength analysis, a multi-level method has been implemented that combines classical complexity assessment criteria with modern approaches to machine analysis. In particular, the length, character composition, pattern detection (repetition, sequences, common keyboard combinations), checking for the presence of known merged password databases, as well as lexical and semantic analysis are carried out. Special attention is paid to the machine approach to weak password detection, which is built on the bidirectional long-term memory model (BiLSTM). The model is trained on a dataset of passwords containing various names and words and demonstrates accuracy of over 92% accuracy on test samples, using the F1, Precision and Recall metrics. The developed application is implemented in the Python environment using the TensorFlow, NumPy, secrets and other libraries. Comprehensive testing of the developed methods was carried out, including both functional and load scenarios. The compliance of the analysis results with the expected cybersecurity standards (NIST, OWASP, ISO/IEC 27001) was assessed and the effectiveness of the approach was confirmed compared to existing analogues. The results of the work can be used to increase the level of account protection in personal and corporate environments, as well as for implementation in password management systems. The application can become the basis for further research and improvement in the field of adaptive authentication, in particular by integrating behavioral analysis or biometric factors. The complexity of the approach, the use of deep learning models and the possibility of personalization make the developed product a valuable contribution to the field of practical cybersecurity. [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 22235744 |
| DOI: | 10.15276/imms.v15.no2.218 |
