Abstract
Text password has lengthy been dominant user authentication approach used by means of a giant quantity of Internet services. The problem of creating passwords in traditional textual method has a set of problems including using common words, using guessable words, and perhaps even using the same word for more than one site. The main contribution of this work is automatic password generator depends on biometric features that can be extracted from human emotions via speech. In latest years, countless emotional speech datasets in one-of-a-kind languages have been collected; however, Arabic is now not among the languages that have been investigated in the context of emotion recognition. For this purpose, a new Arabic emotional speech dataset, which includes 300 samples for six emotional states (Anger, Disgust, Fear, Happiness, Neutral and Sadness), was constructed. Feature extraction suggested by mixing of spectral and prosodic features simultaneously which improve the performance of the proposed system. Classification in proposed achieved using Support Vector Machine (SVM) technique. Finally, Automatic password generated depends on statistical parameters such as (Min, Max, Mean, Median, Mode, Standard deviation and Variance). Experimental results have proven that proposed system achieved 95% recognition rate on average. Proposed generator enables the generation of passwords that meet real-world requirements, including forced password changes and generation of passwords meeting site-specific requirements.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have