In the modern digital era, everyone is partially or fully integrated with cloud computing to access numerous cloud models, services, and applications. Multi-cloud is a blend of a well-known cloud model under a single umbrella to accomplish all the distinct nature and realm requirements under one service level agreement (SLA). In current era of cloud paradigm as the flood of services, applications, and data access rise over the Internet, the lack of confidentiality of the end user’s credentials is rising to an alarming level. Users typically need to authenticate multiple times to get authority and access the desired services or applications. In this research, we have proposed a completely secure scheme to mitigate multiple authentications usually required from a particular user. In the proposed model, a federated trust is created between two different domains: consumer and provider. All traffic coming towards the service provider is further divided into three phases based on the concerned user’s data risks. Single sign-on (SSO) and multifactor authentication (MFA) are deployed to get authentication, authorization, accountability, and availability (AAAA) to ensure the security and confidentiality of the end user’s credentials. The proposed solution exploits the finding that MFA achieves a better AAAA pattern as compared to SSO.

KSII Transactions on Internet and Information Systems TIIS Paper Details

Traditional image retrieval methods often face challenges in adapting to varying user preferences and dynamic datasets. To address these limitations, this research introduces a novel image retrieval framework utilizing deep deterministic policy gradients (DDPG) augmented with a self-adaptive reward mechanism (SARM). The DDPG-SARM framework dynamically adjusts rewards based on user feedback and retrieval context, enhancing the learning efficiency and retrieval accuracy of the agent. Key innovations include dynamic reward adjustment based on user feedback, context-aware reward structuring that considers the specific characteristics of each retrieval task, and an adaptive learning rate strategy to ensure robust and efficient model convergence. Extensive experimentation with the three distinct datasets demonstrates that the proposed framework significantly outperforms traditional methods, achieving the highest retrieval accuracy having 3.38%, 5.26%, and 0.21% improvement overall as compared to the mainstream models over DermaMNIST, PneumoniaMNIST, and OrganMNIST datasets, respectively. The findings contribute to the advancement of reinforcement learning applications in image retrieval, providing a user-centric solution adaptable to various dynamic environments. The proposed method also offers a promising direction for future developments in intelligent image retrieval systems.

In this era of technology, the ways to share information among people are completely changed. Technology made it fast to connect with people and share their views within second by using different social sites. The need for hours is how to secure data in a more precise form that can survive by any unauthorized users' raids such as copy, changing, insertion, and deletion of data. These types of attacks destroyed the originality of data. Data may be in any contents shape audio, video, and text but text contents are most difficult to survive from unauthorized access and use. As the content of the text is sharing in soft form over different social sites so the chances of attacks are increasing and the value of the text is difficult to maintain. The purposed solution of securing information or text from unauthorized attacks is text watermarking. Watermarking is the most host topic due to increasing attacks and as the technology is growing. To secure confidential information is one of the necessities of this era for preserving the originality and authentication of contents. In recent years different techniques of text watermarking are using for the security purposes of text. Different techniques and algorithms are used but the main requirements of text watermarking are not successfully achieved. These techniques may not give full access to attackers but the value of the text is degraded. This thesis presents a more enhanced technique to maintain the originality of the text.