Existing news recommendation systems often overlook the diversity of recommended content and exhibit popularity bias, resulting in suboptimal performance. To address this issue, this paper introduces a novel news recommendation approach, Popularity- and Position-Aware Contrastive Learning for Retrieval-Driven News Recommendation (PP4RNR). It consists of two modules: Entity-Level Retrieval Augmentation (ERA) and Popularity- and Position-Aware Contrastive Learning (PPCL). The ERA module utilizes both entities and titles to retrieve relevant news. Subsequently, retrieval-augmented news is fused with candidate news using our innovative cascaded attention network, leading to richer and more diverse news semantics. The PPCL module introduces perturbations in the news representation using a Gaussian perturbation vector based on the popularity and position information and then employs contrastive learning to regularize the representation space. Hence, this approach not only deepens the understanding of content diversity but also implicitly mitigates the popularity bias prevalent in current models. Rigorous testing on benchmark datasets demonstrates that our method significantly outperforms a range of state-of-the-art techniques.

The demand for information security is growing with the changes of the times, and the authentication system is an important gateway to ensure information security. Password authentication is the most commonly used authentication method in modern network. However, because the password is easy to be cracked, we need to pay more attention to more authentication methods. In many authentications, the advantage of keystrokes and mouse authentication are more obvious; however, when researchers use mouse dynamics to authenticate, classifier training always require a large amount of data, and when the data is less, there may be inaccurate results. In this paper, a decision-level fusion method of the two classifiers is proposed, which reduces the strong dependence on data during training. In this method, the support vector machine optimized by genetic algorithm and k-nearest-neighbor algorithm are combined to get a lower error rate, which is lower than the error rate generated by the two methods alone.

As a new intelligent technology of identity recognition., face recognition is an important means of security authentication. However, in real life, the face may be blocked by hats, masks and other occlusions, resulting in the existing face recognition algorithms and methods cannot be applied in practical applications. In order to solve the above problems, this paper proposes a partial occlusion face recognition method based on infrared thermal imaging to locate acupoints. Firstly, the face image is collected and the face features are obtained by Principal component analysis (PCA) algorithm. Secondly, the infrared thermal imaging technology is introduced adaptively by judging the occlusion status of the face, and then the individual facial acupoints are located by BP neural network algorithm to assist in the effective recognition of partially occluded faces. The results of simulation show that this recognition method proposed can increase the accuracy of partial occlusion face recognition, which has practical value.

Time synchronization is one of the most fundamental services for wireless sensor networks. Prior studies have investigated the clock stability due to environmental dynamics. In this paper, we demonstrate by experiment that in spite of the surrounding environment, time synchronization is unavoidably impacted by in-network energy heterogeneity, which may incur up to 30-40 ppm clock uncertainty. We mathematically analyze the root cause of such clock uncertainty and propose a protocol called EATS. Sensor nodes with EATS can intelligently select the best synchronization parents that minimize the negative impact of the energy heterogeneity. The selection is robust to multiple impacting factors in the network and provides fine-grained synchronization accuracy. In addition, nodes can make use of local energy information and further calibrate the clocks. In light of this, the logic time maintained among different nodes is more consistent and the synchronization can be performed with a longer re-synchronization interval and less energy consumption. We implement EATS with TelosB motes and evaluate the effectiveness and efficiency of our design through extensive experiments and simulations.

In this paper, we propose a novel clock calibration approach called FLIGHT, which leverages the fact that the fluorescent light intensity changes with a stable period that equals half of the alternating current's. By tuning to the light emitted from indoor fluorescent lamps, FLIGHT can intelligently extract the light period information and achieve network wide time calibration by referring to such a common time reference. The light period can be also viewed as an indoor context indicator. As sampling the light sensor consumes substantially less energy, FLIGHT provides us a lightweight clock calibration and time synchronization solution. In addition, FLIGHT suits various mobility-enabled scenarios and it can work well even when the network is temporarily disconnected. We address a series of practical challenges and implement FLIGHT in TelosB motes. We conduct comprehensive experiments using a 12-node test-bed in both static and mobile environments. Over one-week measurement suggests that compared with existing technologies, FLIGHT can achieve tightly synchronized time with low energy consumption. We further leverage the periodical pattern and upgrade FLIGHT to recognize the ambient indoor/outdoor context, based on which the on/off states of a variety of location-based services can be controlled automatically for mobile devices.

In this paper, we propose a novel clock calibration approach called FLIGHT, which leverages the fact that the fluorescent light intensity changes with a stable period that equals half of the alternating current's. By tuning to the light emitted from indoor fluorescent lamps, FLIGHT can intelligently extract the light period information and achieve network wide time calibration by referring to such a common time reference. We address a series of practical challenges and implement FLIGHT in TelosB motes. We conduct comprehensive experiments using a 12-node test-bed in both static and mobile environments. Over one-week measurement suggests that compared with existing technologies, FLIGHT can achieve tightly synchronized time with low energy consumption.

In this demo, we propose a novel clock calibration approach called FLIGHT, which leverages the fact that the fluorescent light intensity changes with a stable period that equals half of the alternating current's. By tuning to the light emitted from indoor fluorescent lamps, FLIGHT can intelligently extract the light period information and achieve network wide time calibration by referring to such a common time reference. We address a series of practical challenges and implement FLIGHT in TelosB motes. In this demonstration, we will show that by taking advantage of the stability of the AC frequency, the detected light intensity, even from different lamps, exhibits a consistent and stable period. FLIGHT can achieve tightly synchronized time with low energy consumption. In addition, since FLIGHT is independent to the network message exchange, time synchronization can be retained even when the network is temporarily disconnected. Such characteristics particularly suit various mobility-enabled scenarios.