Music recommendation algorithms based on knowledge graph and multi-task feature learning.

During music recommendation scenarios, sparsity and cold start problems are inevitable. Auxiliary information has been utilized in music recommendation algorithms to provide users with more accurate music recommendation results. This study proposes an end-to-end framework, MMSS_MKR, that uses a knowledge graph as a source of auxiliary information to serve the information obtained from it to the recommendation module. The framework exploits Cross & Compression Units to bridge the knowledge graph embedding task with recommendation task modules. We can obtain more realistic triple information and exclude false triple information as much as possible, because our model obtains triple information through the music knowledge graph, and the information obtained through the recommendation module is used to determine the truth of the triple information; thus, the knowledge graph embedding task is used to perform the recommendation task. In the recommendation module, multiple predictions are adopted to predict the recommendation accuracy. In the knowledge graph embedding module, multiple calculations are used to calculate the score. Finally, the loss function of the model is improved to help us to obtain more useful information for music recommendations. The MMSS_MKR model achieved significant improvements in music recommendations compared with many existing recommendation models.

Hypertensive Pressure Mechanosensing Alone Triggers Lipid Droplet Accumulation and Transdifferentiation of Vascular Smooth Muscle Cells to Foam Cells.

Arterial Vascular smooth muscle cells (VSMCs) play a central role in the onset and progression of atherosclerosis. Upon exposure to pathological stimuli, they can take on alternative phenotypes that, among others, have been described as macrophage like, or foam cells. VSMC foam cells make up >50% of all arterial foam cells and have been suggested to retain an even higher proportion of the cell stored lipid droplets, further leading to apoptosis, secondary necrosis, and an inflammatory response. However, the mechanism of VSMC foam cell formation is still unclear. Here, it is identified that mechanical stimulation through hypertensive pressure alone is sufficient for the phenotypic switch. Hyperspectral stimulated Raman scattering imaging demonstrates rapid lipid droplet formation and changes to lipid metabolism and changes are confirmed in ABCA1, KLF4, LDLR, and CD68 expression, cell proliferation, and migration. Further, a mechanosignaling route is identified involving Piezo1, phospholipid, and arachidonic acid signaling, as well as epigenetic regulation, whereby CUT&Tag epigenomic analysis confirms changes in the cells (lipid) metabolism and atherosclerotic pathways. Overall, the results show for the first time that VSMC foam cell formation can be triggered by mechanical stimulation alone, suggesting modulation of mechanosignaling can be harnessed as potential therapeutic strategy.

Real-time low-complexity diversity combining algorithm for free space coherent optical communication systems over atmospheric turbulence channel.

A novel diversity combining scheme, in conjunction with the complex-valued decision-directed least mean square (CV-DD-LMS) algorithm, is evaluated, and a real-time experimental validation is presented. This proposed scheme employs the CV-DD-LMS algorithm to concurrently perform beam combination and carrier phase recovery (CPR), thereby effectively reducing the overall complexity of digital signal processing. Furthermore, in the numerical simulation, under a low signal-to-noise ratio (SNR), a scheme utilizing the CV-DD-LMS algorithm effectively avoids cycle slips (CS) and outperforms schemes employing independent CPR modules. We experimentally validate this novel scheme by implementing it on an FPGA in a real-time 2.5Gb/s QPSK diversity-receiving system with three inputs. The back-to-back sensitivity is assessed using static received optical power, while the dynamic performance is evaluated by employing variable optical attenuators (VOAs) to simulate a power fluctuation at a frequency of 100kHz. The result proves that the implementation of the CV-DD-LMS algorithm yields stable performance while effectively reducing computational complexity.

Nonlinear transmission performance comparison employing 60 × 416.7-Gb/s TPS-64QAM and UD-16QAM systems with limited bandwidth.

The impacts of limited bandwidth on the nonlinear transmission performance are investigated by employing a truncated probabilistic shaped 64-ary quadrature amplitude modulation (TPS-64QAM) and a uniformly distributed 16-ary quadrature amplitude modulation (UD-16QAM) over a bandwidth-limited 75-GHz spaced 25-Tb/s (60 × 416.7 Gb/s) 6300-km transmission system. In terms of nonlinear performance measured by optimal launch power, theoretical analyses show that a 0.4-dB improvement could be introduced by UD-16QAM with respect to TPS-64QAM over a 6300-km transmission without limited bandwidth. However, contrary results would be observed that TPS-64QAM would outperform UD-16QAM by about 0.8 dB in terms of optimal launch power when the impacts of limited bandwidth are considered. Besides, numerical simulations and experimental results could both validate that about 1.0-dB optimal launch power improvement could be obtained by TPS-64QAM under bandwidth-limited cases, which is roughly similar to the results of theoretical analyses. Additionally, WDM experimental results show that all 60 tested channels could agree with the BER requirements by employing TPS-64QAM, further validating the superiority of TPS-64QAM compared to UD-16QAM under bandwidth-limited cases.

Causal Role for Neutrophil Elastase in Thoracic Aortic Dissection in Mice.

BACKGROUND: Thoracic aortic dissection (TAD) is a life-threatening aortic disease without effective medical treatment. Increasing evidence has suggested a role for NE (neutrophil elastase) in vascular diseases. In this study, we aimed at investigating a causal role for NE in TAD and exploring the molecular mechanisms involved. METHODS: ß-aminopropionitrile monofumarate was administrated in mice to induce TAD. NE deficiency mice, pharmacological inhibitor GW311616A, and adeno-associated virus-2-mediated in vivo gene transfer were applied to explore a causal role for NE and associated target gene in TAD formation. Multiple functional assays and biochemical analyses were conducted to unravel the underlying cellular and molecular mechanisms of NE in TAD. RESULTS: NE aortic gene expression and plasma activity was significantly increased during ß-aminopropionitrile monofumarate-induced TAD and in patients with acute TAD. NE deficiency prevents ß-aminopropionitrile monofumarate-induced TAD onset/development, and GW311616A administration ameliorated TAD formation/progression. Decreased levels of neutrophil extracellular traps, inflammatory cells, and MMP (matrix metalloproteinase)-2/9 were observed in NE-deficient mice. TBL1x (F-box-like/WD repeat-containing protein TBL1x) has been identified as a novel substrate and functional downstream target of NE in TAD. Loss-of-function studies revealed that NE mediated inflammatory cell transendothelial migration by modulating TBL1x-LTA4H (leukotriene A4 hydrolase) signaling and that NE regulated smooth muscle cell phenotype modulation under TAD pathological condition by regulating TBL1x-MECP2 (methyl CpG-binding protein 2) signal axis. Further mechanistic studies showed that TBL1x inhibition decreased the binding of TBL1x and HDAC3 (histone deacetylase 3) to MECP2 and LTA4H gene promoters, respectively. Finally, adeno-associated virus-2-mediated Tbl1x gene knockdown in aortic smooth muscle cells confirmed a regulatory role for TBL1x in NE-mediated TAD formation. CONCLUSIONS: We unravel a critical role of NE and its target TBL1x in regulating inflammatory cell migration and smooth muscle cell phenotype modulation in the context of TAD. Our findings suggest that the NE-TBL1x signal axis represents a valuable therapeutic for treating high-risk TAD patients.

Targeted Inhibition of Matrix Metalloproteinase-8 Prevents Aortic Dissection in a Murine Model.

Aortic dissection (AD) is a lethal aortic pathology without effective medical treatments since the underlying pathological mechanisms responsible for AD remain elusive. Matrix metalloproteinase-8 (MMP8) has been previously identified as a key player in atherosclerosis and arterial remodeling. However, the functional role of MMP8 in AD remains largely unknown. Here, we report that an increased level of MMP8 was observed in 3-aminopropionitrile fumarate (BAPN)-induced murine AD. AD incidence and aortic elastin fragmentation were markedly reduced in MMP8-knockout mice. Importantly, pharmacologic inhibition of MMP8 significantly reduced the AD incidence and aortic elastin fragmentation. We observed less inflammatory cell accumulation, a lower level of aortic inflammation, and decreased smooth muscle cell (SMC) apoptosis in MMP8-knockout mice. In line with our previous observation that MMP8 cleaves Ang I to generate Ang II, BAPN-treated MMP8-knockout mice had increased levels of Ang I, but decreased levels of Ang II and lower blood pressure. Additionally, we observed a decreased expression level of vascular cell adhesion molecule-1 (VCAM1) and a reduced level of reactive oxygen species (ROS) in MMP8-knockout aortas. Mechanistically, our data show that the Ang II/VCAM1 signal axis is responsible for MMP8-mediated inflammatory cell invasion and transendothelial migration, while MMP8-mediated SMC inflammation and apoptosis are attributed to Ang II/ROS signaling. Finally, we observed higher levels of aortic and serum MMP8 in patients with AD. We therefore provide new insights into the molecular mechanisms underlying AD and identify MMP8 as a potential therapeutic target for this life-threatening aortic disease.

Distributed measurement of mode group effective refractive index difference in a few mode optical fibers.

The possibility to perform distributed measurements of the effective refractive index difference between distinct modes in few mode optical fibers is demonstrated using phase sensitive optical time domain reflectometry. Effective refractive index differences between LP02, LP21a and LP21b modes are measured with for a spatial resolution of 24m.

Enlarging the frequency dynamic range of DMZI sensors based on the switching control feedback system.

The feedback loop in dual Mach-Zehnder interferometer (DMZI) sensors stabilizes the system operating at the quadratic point for the highest sensitivity but requires the minimum measurable vibration frequency out of the feedback bandwidth, resulting in a limited dynamic range. In this paper, we point out that the feedback operation is unnecessary while vibration is occurring and propose a strategy to adaptively enable/disable the feedback phase compensation depending on the vibration state, lowering the minimum measurable vibration frequency tenfold. Moreover, the state variable employed enables direct extraction of vibration-related data, with no need of complicated postprocessing algorithms.

Closed-loop technique based on gain balancing for real-time Brillouin optical time-domain analysis.

A closed-loop servo control based on balancing the gain of two probing frequencies is proposed for real-time Brillouin optical time-domain analysis (BOTDA) without post-processing. With the most basic BOTDA hardware setup, the system can perform measurement in 150 ms and track a sudden Brillouin frequency shift (BFS) change in excess of 300 MHz (corresponding to a temperature change of more than 250°C) over ∼5 km of fiber with a spatial resolution of 2 m. Moreover, the feedback loop is independent of the loss experienced by the probe and pump, with no requirement on the BFS uniformity along the fiber. All these advantages make the proposed system suitable for field applications in harsh environments.

Impact of optical noises on unipolar-coded Brillouin optical time-domain analyzers.

Noise models for both single-pulse and coded Brillouin optical time-domain analyzers (BOTDA) are established to quantify the actual signal-to-noise ratio (SNR) enhancement provided by pulse coding at any fiber position and in any operating condition. Simulation and experimental results show that the polarization noise and spontaneous Brillouin scattering (SpBS) to signal beating noise could highly penalize the performance of coded-BOTDA, depending on the code type and the interrogated fiber position. The models also serve as a useful tool to optimize the SNR improvement by trading off the accumulated Brillouin gain and optical noises.

Long-distance distributed pressure sensing based on frequency-scanned phase-sensitive optical time-domain reflectometry.

In this paper, a long-distance distributed pressure sensing system based on a special fiber and using frequency-scanned phase-sensitive optical time-domain reflectometry is proposed. The fiber shows high pressure sensitivity (159 MHz/bar) and low loss (3 dB/km) owing to its simple structure made of two large air holes in the cladding. The pressure response of the two orthogonal polarization axes of the fiber is explored distinctively. Distributed pressure sensing over a long sensing range (720 m) and high spatial resolution (5 cm) is demonstrated, resulting in 14,400 resolved sensing points with uncertainty on pressure of 0.49 bar. Discrimination between the temperature/strain and pressure responses is demonstrated, taking advantage of the different pressure and temperature sensitivities of the two polarization axes. In addition, the temperature response of the fiber is studied and the simulation results show the possibility of scaling the temperature sensitivity by adjusting the size of the core. The sensing distance limit due to crosstalk between the polarization axes is also discussed.

Recommendation algorithm based on attributed multiplex heterogeneous network.

In the field of deep learning, the processing of large network models on billions or even tens of billions of nodes and numerous edge types is still flawed, and the accuracy of recommendations is greatly compromised when large network embeddings are applied to recommendation systems. To solve the problem of inaccurate recommendations caused by processing deficiencies in large networks, this paper combines the attributed multiplex heterogeneous network with the attention mechanism that introduces the softsign and sigmoid function characteristics and derives a new framework SSN_GATNE-T (S represents the softsign function, SN represents the attention mechanism introduced by the Softsign function, and GATNE-T represents the transductive embeddings learning for attribute multiple heterogeneous networks). The attributed multiplex heterogeneous network can help obtain more user-item information with more attributes. No matter how many nodes and types are included in the model, our model can handle it well, and the improved attention mechanism can help annotations to obtain more useful information via a combination of the two. This can help to mine more potential information to improve the recommendation effect; in addition, the application of the softsign function in the fully connected layer of the model can better reduce the loss of potential user information, which can be used for accurate recommendation by the model. Using the Adam optimizer to optimize the model can not only make our model converge faster, but it is also very helpful for model tuning. The proposed framework SSN_GATNE-T was tested for two different types of datasets, Amazon and YouTube, using three evaluation indices, ROC-AUC (receiver operating characteristic-area under curve), PR-AUC (precision recall-area under curve) and F1 (F1-score), and found that SSN_GATNE-T improved on all three evaluation indices compared to the mainstream recommendation models currently in existence. This not only demonstrates that the framework can deal well with the shortcomings of obtaining accurate interaction information due to the presence of a large number of nodes and edge types of the embedding of large network models, but also demonstrates the effectiveness of addressing the shortcomings of large networks to improve recommendation performance. In addition, the model is also a good solution to the cold start problem.

Genetic-optimised aperiodic code for distributed optical fibre sensors.

Distributed optical fibre sensors deliver a map of a physical quantity along an optical fibre, providing a unique solution for health monitoring of targeted structures. Considerable developments over recent years have pushed conventional distributed sensors towards their ultimate performance, while any significant improvement demands a substantial hardware overhead. Here, a technique is proposed, encoding the interrogating light signal by a single-sequence aperiodic code and spatially resolving the fibre information through a fast post-processing. The code sequence is once forever computed by a specifically developed genetic algorithm, enabling a performance enhancement using an unmodified conventional configuration for the sensor. The proposed approach is experimentally demonstrated in Brillouin and Raman based sensors, both outperforming the state-of-the-art. This methodological breakthrough can be readily implemented in existing instruments by only modifying the software, offering a simple and cost-effective upgrade towards higher performance for distributed fibre sensing.

Distributed and dynamic strain sensing with high spatial resolution and large measurable strain range.

A distributed and dynamic strain sensing system based on frequency-scanning phase-sensitive optical time domain reflectometry is proposed and demonstrated. By utilizing an RF pulse scheme with a fast arbitrary waveform generator, a train of optical pulses covering a large range of different optical frequencies, short pulse width, and high extinction ratio is generated. Also, a Rayleigh-enhanced fiber is used to eliminate the need for averaging, allowing single-shot operation. Using direct detection and harnessing a dedicated least mean square algorithm, the method exhibits a record high spatial resolution of 20 cm, concurrently with a large measurable strain range (80µÎµ, 60 demonstrated), a fast sampling rate of 27.8 kHz (almost the maximum possible for a 55 m long fiber and 60 frequency steps), and low strain noise floor (<1.8nε/Hz for vibrations below 700 Hz and <0.7nε/Hz for higher frequencies).

Short spatial resolution retrieval from a long pulse Brillouin optical time-domain analysis trace.

A novel, to the best of our knowledge, postprocessing technique is proposed to extract with a flexible and variable spatial resolution the information from Brillouin optical time-domain analyzers, obtained using a pulse longer than the acoustic settling time. The negative impact of the acoustic transient effect is suppressed, enabling a Brillouin response proportional to the spatial resolution and a Brillouin gain spectrum keeping its natural linewidth. This leads to a better overall sensing performance, in particular for submetric spatial resolutions, with no compromises on sensing range and measurement time.

Study on the signal-to-noise ratio of Brillouin optical-time domain analyzers.

The signal-to-noise ratio (SNR) of Brillouin optical time-domain analyzers (BOTDA) is modelled and experimentally validated, using direct detection with and without the use of optical pre-amplification. The behavior of SNR as a function of the Brillouin gain and the probe power reaching the photo detection is analyzed in depth using this developed model and checked using two photodetectors with different specifications. It proves that a pre-amplification associated to a good-quality photodetector and a well-matched post-processing filtering can secure the highest SNR for direct-detection BOTDA. Such an optimal SNR presents only a 2.3 dB penalty compared to the ideal shot-noise-limited case that can only be reached using rather sophisticated configurations. In addition, the model here established predicts the SNR at any fiber position in any given experimental condition.

High-resolution distributed shape sensing using phase-sensitive optical time-domain reflectometry and multicore fibers.

In this paper, a highly-sensitive distributed shape sensor based on a multicore fiber (MCF) and phase-sensitive optical time-domain reflectometry (φ-OTDR) is proposed and experimentally demonstrated. The implemented system features a high strain sensitivity (down to ∼0.3 µÉ›) over a 24 m-long MCF with a spatial resolution of 10 cm. The results demonstrate good repeatability of the relative fiber curvature and bend orientation measurements. Changes in the fiber shape are successfully retrieved, showing detectable displacements of the free moving fiber end as small as 50 µm over a 60 cm-long fiber. In addition, the proposed technique overcomes cross-sensitivity issues between strain and temperature. To the best of our knowledge, the results presented in this work provide the first demonstration of distributed shape sensing based on φ-OTDR using MCFs. This high-sensitivity technique proves to be a promising approach for a wide range of new applications such as dynamic, long distance and three-dimensional distributed shape sensing.

Hybrid Golay-coded Brillouin optical time-domain analysis based on differential pulses.

Different approaches to implement unipolar Golay coding in Brillouin optical time-domain analysis based on a differential pulse pair (DPP) are investigated. The analysis points out that dedicated post-processing procedures must be followed to secure the sharp spatial resolution associated with the DPP method. Moreover, a novel hybrid Golay-DPP coding scheme is proposed, offering 1.5 dB signal-to-noise ratio improvement with respect to traditional unipolar Golay coding, while halving the measurement time, constituting a 3 dB overall coding gain enhancement. Proof-of-concept experiments validate the proposed technique, demonstrating a 50 cm spatial resolution over a 10.164 km long sensing fiber with a frequency uncertainty of 1.4 MHz.