In response to the rapid growth of credit-investigation data, data redundancy among credit-investigation agencies, privacy leakages of credit-investigation data subjects, and data security risks have been reported. This study proposes a privacy-protection scheme for a credit-investigation system based on blockchain technology, which realizes the secure sharing of credit-investigation data among multiple entities such as credit-investigation users, credit-investigation agencies, and cloud service providers. This scheme is based on blockchain technology to solve the problem of islanding of credit-investigation data and is based on zero-knowledge-proof technology, which works by submitting a proof to the smart contract to achieve anonymous identity authentication, ensuring that the identity privacy of credit-investigation users is not disclosed; this scheme is also based on searchable-symmetric-encryption technology to realize the retrieval of the ciphertext of the credit-investigation data. A security analysis showed that this scheme guarantees the confidentiality, the availability, the tamper-proofability, and the ciphertext searchability of credit-investigation data, as well as the fairness and anonymity of identity authentication in the credit-investigation data query. An efficiency analysis showed that, compared with similar identity-authentication schemes, the proof key of this scheme is smaller, and the verification time is shorter. Compared with similar ciphertext-retrieval schemes, the time for this scheme to generate indexes and trapdoors and return search results is significantly shorter.

Electronic health record (EHR) cloud system, as a primary tool driving the informatization of medical data, have positively impacted both doctors and patients by providing accurate and complete patient information. However, ensuring the security of EHR cloud system remains a critical issue. Some patients require regular remote medical services, and controlling access to medical data involving patient privacy during specific times is essential. Timed-release encryption (TRE) technology enables the sender to preset a future time T at which the data can be decrypted and accessed. It is a cryptographic primitive with time-dependent properties. Currently, mainstream TRE schemes are based on non-interactive single time server methods. However, if the single time server is attacked or corrupted, it is easy to directly threaten the security applications of TRE. Although some research schemes “distribute” the single time server into multiple ones, they still cannot resist the single point of failure problem. To address this issue, we propose a multiple time servers TRE scheme based on Shamir secret sharing and another variant derived from it. In our proposed schemes, the data receiver does not need to interact with the time servers; instead, they only need to obtain the time trapdoors that exceed or equal the preset threshold value for decryption, which ensures the identity privacy of the data sender and tolerates partial downtime or other failures of some time servers, significantly improving TRE reliability. Security analysis indicates that our proposed schemes demonstrate data confidentiality, verifiability, anti-advance decryption, and robust decryption with multiple time trapdoors, making them more practical. Efficiency analysis indicates that although our schemes have slightly higher computational costs than most efficient existing TRE schemes, such differences are insignificant from a practical application perspective.

Feature selection in high-dimensional data is an important part of the data mining process and is widely used in bioinformatics, statistics and image processing fields. Successfully selecting informative features can significantly improve learning accuracy and improve result comprehensibility. However, it is a challenging problem to select features accurately and efficiently from high-dimensional data. In this paper, we propose a Weighted Sparse Regression with Mutual Information (WSRMI) for selecting structural features. Differing from traditional sparse feature selection models that focus solely on either feature correlations or feature importance, the proposed model integrates both aspects through a mutual-information-based weighting mechanism. The proposed model can be effectively applied to regression and binary classification tasks, making it more general and practical for real-world applications. The proposed model is statistically compared with several existing classical models over randomly generated classification and benchmark datasets. Experimental results show that the proposed model is more effective at selecting the informative features with a superior prediction performance than the comparative ones.

Federated learning (FL) has gained significant attention as a distributed machine learning approach in recent years. However, in the context of the Internet of Things, terminal devices exhibit considerable heterogeneity in data distribution, communication, and computational resources. This heterogeneity can introduce bias, hindering model accuracy through iterative training. Moreover, disparities in communication and computational capabilities limit FL to resource-constrained devices, slowing convergence. To address these challenges, we propose FedDG, an FL framework that integrates model compression and dynamic fusion. Specifically, for nonindependently and identically distributed data, we adaptively adjust the client compression rate based on model training errors, reducing transmission costs while improving both convergence speed and accuracy. Furthermore, clients are dynamically selected for aggregation in each round based on training time and model performance, alleviating the impact of low-resource clients and reducing communication overhead. Experimental results show that FedDG significantly lowers communication bandwidth and data transmission size, and accelerates model convergence under data heterogeneity compared to benchmark schemes, such as FedAvg, FedNova, STC, DF_FL, DAdaQuant, and adaptive quantization in device selection strategy (AQUILA).