Search

Transformer models have had a great impact on natural language processing (NLP) in recent years by realizing outstanding and efficient contextualized language models. Recent studies have used transformer‐based language models for various NLP tasks, including Persian named entity recognition (NER). However, in complex tasks, for example, NER, it is difficult to determine which contextualized embedding will produce the best representation for the tasks. Considering the lack of comparative studies to investigate the use of different contextualized pretrained models with sequence modeling classifiers, we conducted a comparative study about using different classifiers and embedding models. In this paper, we use different transformer‐based language models tuned with different classifiers, and we evaluate these models on the Persian NER task. We perform a comparative analysis to assess the impact of text representation and text classification methods on Persian NER performance. We train and evaluate the models on three different Persian NER datasets, that is, MoNa, Peyma, and Arman. Experimental results demonstrate that XLM‐R with a linear layer and conditional random field (CRF) layer exhibited the best performance. This model achieved phrase‐based F‐measures of 70.04, 86.37, and 79.25 and word‐based F scores of 78, 84.02, and 89.73 on the MoNa, Peyma, and Arman datasets, respectively. These results represent state‐of‐the‐art performance on the Persian NER task. [ABSTRACT FROM AUTHOR]

The network traffic prediction of a smart substation is key in strengthening its system security protection. To improve the performance of its traffic prediction, in this paper, we propose an improved gravitational search algorithm (IGSA), then introduce the IGSA into a wavelet neural network (WNN), iteratively optimize the initial connection weighting, scalability factor, and shift factor, and establish a smart substation network traffic prediction model based on the IGSA‐WNN. A comparative analysis of the experimental results shows that the performance of the IGSA‐WNN‐based prediction model further improves the convergence velocity and prediction accuracy, and that the proposed model solves the deficiency issues of the original WNN, such as slow convergence velocity and ease of falling into a locally optimal solution; thus, it is a better smart substation network traffic prediction model. [ABSTRACT FROM AUTHOR]

This paper proposes a new min-sum algorithm for lowdensity parity-check decoding. In this paper, we first define the negative and positive effects of the received signal-to-noise ratio (SNR) in the min-sum decoding algorithm. To improve the performance of error correction by considering the negative and positive effects of the received SNR, the proposed algorithm applies adaptive scaling factors not only to extrinsic information but also to a received log-likelihood ratio. We also propose a combined variable and check node architecture to realize the proposed algorithm with low complexity. The simulation results show that the proposed algorithm achieves up to 0.4 dB coding gain with low complexity compared to existing min-sum-based algorithms. [ABSTRACT FROM AUTHOR]

An integrated multi-beam satellite and multi-cell terrestrial system is an attractive means for highly efficient communication due to the fact that the two components (satellite and terrestrial) make the most of each other's resources. In this paper, a terrestrial component reuses a satellite's resources under the control of the satellite's network management system. This allows the resource allocation for the satellite and terrestrial components to be coordinated to optimize spectral efficiency and increase overall system capacity. In such a system, the satellite resources reused in the terrestrial component may bring about severe interference, which is one of the main factors affecting system capacity. Under this consideration, the objective of this paper is to achieve an optimized resource allocation in both components in such a way as to minimize any resulting inter-component interference. The objective of the proposed scheme is to mitigate this inter-component interference by optimizing the total transmission power -- the result of which can lead to an increase in capacity. The simulation results in this paper illustrate that the proposed scheme affords a more energy-efficient system to be implemented, compared to a conventional power management scheme, by allocating the bandwidth uniformly regardless of the amount of interference or traffic demand. [ABSTRACT FROM AUTHOR]

Background: There is controversy regarding the best technique for applying Kinesio Taping (KT), and the theory supporting that skin convolutions may explain its efficacy has recently been challenged.Objective: To compare the immediate and short-term effectiveness of KT tightness on mechanosensitivity and spinal mobility in nonspecific low back pain (LBP), and to observe the influence of gender in the outcome measures.Design: Double-blind, randomized, controlled trial.Setting: University-based clinical research center.Participants: A total of 75 individuals with a mean age of 33 years (±7.4 years), 60% female and 40% male, with nonspecific LBP were recruited and randomly assigned to 1 of the following study groups: standard KT tension (n = 26), increased KT tension (n = 25), and no KT tension (n = 24).Interventions: All participants received a two I-strip taping over the paravertebral muscles for 24 hours. Paper-off tension (15%-25% of the available stretch) was used in the standard KT group, which was increased to 40% in the increased KT tension group. The rest of participants received a taping procedure with no KT tension. Measurements were taken at baseline, immediately after the taping, 24 hours after the taping, and after KT removal.Main Outcome Measures: The primary outcome included pressure pain thresholds over the erector spinae and gluteus medius muscles. The secondary outcome was lumbar mobility (assessed with a digital inclinometer, and back-saver sit-and-reach, finger-to-floor, and sit-and-reach tests).Results: In the between-groups analysis of the mean score changes after baseline assessment, no significant differences were found for any of the outcome measures (P > .05) except the left back-saver sit-and-reach test (P = .03). A statistically significant interaction group × gender × time was observed only for mechanosensitivity values (P = .02 for the gluteus and P = .01 for the erector spinae).Conclusion: KT tightness does not seem to influence pain sensitivity and lumbar mobility in chronic LBP in either the immediate or short term.Level Of Evidence: II. [ABSTRACT FROM AUTHOR]

In connection-oriented data transport services, data loss can occur when a service experiences a problem in its end to- end path. To resolve the problem promptly, the data transport systems providing the service must quickly modify their internal configurations, which are distributed among different locations within each system. The configurations are modified through a series of problem (event) handling procedures, which are carried out by multiple control processors in the system. This paper proposes a provisioning-to-signaling method for intercontrol-processor messaging to improve the time efficiency of event processing. This method simplifies the sharing of the runtime event, and minimizes the time variability caused by the amount of event data, which results in a decrease in the latency time and an increase in the time determinacy when processing global events. The proposed method was tested for an event that required 4,000 internal path changes, and was found to lessen the latency time of global event processing by about 50% compared with the time required for general methods to do the same; in addition, it reduced the impact of the event data on the event processing time to about 30%. [ABSTRACT FROM AUTHOR]

In this paper, a new structure for an advanced high holding voltage silicon controlled rectifier (AHHVSCR) is proposed. The proposed new structure specifically for an AHHVSCR-based electrostatic discharge (ESD) protection circuit can protect integrated circuits from ESD stress. The new structure involves the insertion of a PMOS into an AHHVSCR so as to prevent a state of latch-up from occurring due to a low holding voltage. We use a TACD simulation to conduct a comparative analysis of three types of circuit -- (i) an AHHVSCR-based ESD protection circuit having the proposed new structure (that is, a PMOS inserted into the AHHVSCR), (ii) a standard AHHVSCR-based ESD protection circuit, and (iii) a standard HHVSCR-based ESD protection circuit. A circuit having the proposed new structure is fabricated using 0.18 µm Bipolar-CMOS-DMOS technology. The fabricated circuit is also evaluated using Transmission- Line Pulse measurements to confirm its electrical characteristics, and human-body model and machine model tests are used to confirm its robustness. The fabricated circuit has a holding voltage of 18.78 V and a second breakdown current of more than 8 A. [ABSTRACT FROM AUTHOR]

In this paper, we propose a classification-based hybridization method for selecting the best translation results between a statistical machine translation (SMT) and a rule-based machine translation (RBMT). Given the source sentence, a classifier extracts the feature inputs and determines the translation system, usually attaching a label to indicate the system as an output. In learning such a classifier, the generation of the training dataset and feature extraction affect the quality of the hybridization. In previous approaches for the hybridization, the labels are determined based on the quality evaluation metrics such as from the bilingual evaluation understudy (BLEU) or National Institute of Standards and Technology (NIST). This approach has a limit in hybridizing SMT with RBMT, because such measures were designed for evaluating the fluency and SMT has an advantage over RBMT. This may cause a biased preference for SMT. To manage this issue, we propose a method to cut off an uncertain translation from a label prediction. In our experiment, this generation method improved the translation accuracy by 5.0%, compared to labeling through competition. To handle the second issue for extracting features, we investigate better combinations of feature sets divided into six groups. Depending on their use of internal information of MT systems, these groups are either black-box or glass-box groups. We split them again based on the source of the sentence information: the source language side, the target language side, or both. In the experiments, using all of the groups showed the best translation quality and classifier accuracy, as compared with other feature combinations. Overall, the proposed classification-based hybrid approach achieved an 81.5% translation quality, while an RBMT and a phrase-based SMT show a translation quality of 77.1% and 76.3%, respectively. [ABSTRACT FROM AUTHOR]

Until now, a lot of hardware countermeasure logics have been developed, including SABL, TDPL, DyCML and so on. Current mode based logic styles are in use as those consume less power compared to voltage mode based logic styles such as SABL and TDPL. Although we developed TPDyCML in 2012 and presented it at the WISA 2012 conference, we have further optimized it using the BDD algorithm and confirmed the properties through the practical implementation of the AES S-box. In this paper, we will explain the outcome of a HSPICE simulation attack including a correlation power attack to AES S-box with compact NMOS tree using BDD in CMOS, SABL, TDPL, DyCML and TPDyCML. Also, to compare the performance of each logic style in greater detail we will also carry out mutual information analysis (MIA). In the results, we have confirmed that our logic style has good properties and has 15% less information leakage than other secure logic styles. [ABSTRACT FROM AUTHOR]

In this paper, a lateral power metal-oxide- semiconductor field-effect transistor with ultra-low specific on-resistance is proposed to be applied to a highvoltage (up to 200 V) integrated chip. The proposed structure has two characteristics. Firstly, a high level of drift doping concentration can be kept because a tiltimplanted p-drift layer assists in the full depletion of the n-drift region. Secondly, charge imbalance is avoided by an extended trench gate, which suppresses the trench corner effect occurring in the n-drift region and helps achieve a high breakdown voltage (BV). Compared to a conventional trench gate, the simulation result shows a 37.5% decrease in Ron.sp and a 16% improvement in BV. [ABSTRACT FROM AUTHOR]

To solve the increasing popularity of video streaming services over the Internet, recent research activities have addressed the locality of content delivery from a network edge by introducing a storage module into a router. To employ in-network caching and persistent request routing, this paper introduces a hybrid content delivery network (CDN) system combining novel content routers in an underlay together with a traditional CDN server in an overlay. This system first selects the most suitable delivery scheme (that is, multicast or broadcast) for the content in question and then allocates an appropriate number of channels based on a consideration of the content's popularity. The proposed scheme aims to minimize traffic volume and achieve optimal delivery cost, since the most popular content is delivered through broadcast channels and the least popular through multicast channels. The performance of the adaptive scheme is clearly evaluated and compared against both the multicast and broadcast schemes in terms of the optimal in-network caching size and number of unicast channels in a content router to observe the significant impact of our proposed scheme. [ABSTRACT FROM AUTHOR]

The grass tribe Andropogoneae includes plants of great economic importance, such as maize, sugarcane, sorghum, and others. Recent works have questioned the use of the name Andropogoneae Dumort. (1824) for the tribe and have suggested erroneously that the correct name should be Sacchareae Martinov (1820). However, Martinov's name was not published at the rank of tribe. Therefore, it does not threaten the well-established name Andropogoneae. In this paper we lay out the basis for the controversy and the rationale for continued use of the name Andropogoneae. [ABSTRACT FROM AUTHOR]

Research quantifying ecosystem services (ES) - collectively, the benefits that society obtains from ecosystems - is rapidly increasing. Despite the seemingly straightforward definition, a wide variety of methods are used to measure ES. This methodological variability has largely been ignored, and standard protocols to select measures that capture ES provision have yet to be established. Furthermore, most published papers do not include explicit definitions of individual ES. We surveyed the literature on pollination ES to assess the range of measurement approaches, focusing on three essential steps: (1) definition of the ES, (2) identification of components contributing to ES delivery, and (3) selection of metrics to represent these components. We found considerable variation in how pollination as an ES - a relatively well-defined service - is measured. We discuss potential causes of this variability and provide suggestions to address this issue. Consistency in ES measurement, or a clear explanation of selected definitions and metrics, is critical to facilitate comparisons among studies and inform ecosystem management. [ABSTRACT FROM AUTHOR]

In most spectral clustering approaches, the Gaussian kernel-based similarity measure is used to construct the affinity matrix. However, such a similarity measure does not work well on a dataset with a nonlinear and elongated structure. In this paper, we present a new similarity measure to deal with the nonlinearity issue. The maximum flow between data points is computed as the new similarity, which can satisfy the requirement for similarity in the clustering method. Additionally, the new similarity carries the global and local relations between data. We apply it to spectral clustering and compare the proposed similarity measure with other state-of-the-art methods on both synthetic and real-world data. The experiment results show the superiority of the new similarity: 1) The max-flow-based similarity measure can significantly improve the performance of spectral clustering; 2) It is robust and not sensitive to the parameters. [ABSTRACT FROM AUTHOR]

Nowadays, the multicore processor is watched with interest by people all over the world. As the design technology of system on chip has developed, observing and controlling the processor core's internal state has not been easy. Therefore, multicore processor debugging is very difficult and time-consuming. Thus, we need a reliable and efficient debugger to find the bugs. In this paper, we propose an on-chip debug architecture for multicore processors that is easily adaptable and flexible. It is based on the JTAG standard and supports monitoring mode debugging, which is different from run-stop mode debugging. Compared with the debug architecture that supports the run-stop mode debugging, the proposed architecture is easily applied to a debugger and has the advantage of having a desirable gate count and execution cycle. To verify the on-chip debug architecture, it is applied to the debugger of the prototype multicore processor and is tested by interconnecting it with a software debugger based on GDB and configured for the target processor. [ABSTRACT FROM AUTHOR]

With the increasing demands for mobile wireless sensor networks in recent years, designing an energy-efficient clustering and routing protocol has become very important. This paper provides an analytical model to evaluate the power consumption of a mobile sensor node. Based on this, a clustering algorithm is designed to optimize the energy efficiency during cluster head formation. A genetic algorithm technique is employed to find the near-optimal threshold for residual energy below which a node has to give up its role of being the cluster head. This clustering algorithm along with a hybrid routing concept is applied as the near-optimal energy-efficient routing technique to increase the overall efficiency of the network. Compared to the mobile low energy adaptive clustering hierarchy protocol, the simulation studies reveal that the energy-efficient routing technique produces a longer network lifetime and achieves better energy efficiency. [ABSTRACT FROM AUTHOR]

Background: Systemic fluoroquinolones and tetracyclines reach steady-state levels in gingival crevicular fluid (GCF) that are several-fold higher than their levels in serum. The mechanism by which this occurs is unclear, but gingival fibroblasts are known to accumulate these agents. Uptake by fibroblasts could enhance their distribution to gingiva. To test this hypothesis, steady-state levels of these agents were assayed in serum, gingival connective tissue (GCT), and GCF.Methods: Healthy subjects who needed resective periodontal surgery participated in the study. Approximately 78 hours prior to the surgical appointment, each subject began a 3-day regimen of ciprofloxacin or doxycycline. At the surgical appointment (scheduled approximately 6 hours after the last dose), samples of blood and GCT were collected. GCF samples were collected on paper strips and measured with an electronic device. Samples were extracted and analyzed by high performance liquid chromatography.Results: Mean ciprofloxacin levels in serum, GCT, and GCF were 0.40 microg/ml, 1.38 microg/g, and 1.66 microg/ml, respectively (P<0.001, N=9). For doxycycline, these levels were 1.11 microg/ml, 2.03 microg/g, and 2.41 microg/ml, respectively (P=0.002, N=8). For both agents, the GCT and GCF levels were significantly higher than serum levels (P<0.05), but not significantly different from each other.Conclusions: Our findings suggest that fibroblasts could play an important role in the distribution of fluoroquinolones and tetracyclines to the gingiva. By accumulating these agents in GCT, fibroblasts could contribute to the relatively high levels they attain in GCF. [ABSTRACT FROM AUTHOR]

The off-line learning with a supervisor in the recurrent neural network can be considered as the minimum search on the surface (called learning surface) formed by the evaluation functions. The learning surface in the search has a characteristic shape, and the authors are proposing the valley searching method, based on the global shape of the learning surface. As the search for the minimum, various methods, such as the steepest descent (gradient) and the conjugate gradient methods, often are employed. The relation between the search method and the characteristic shape of the learning surface has not been clarified and it has been difficult to assess quantitatively the usefulness of the searching methods. From such a viewpoint, this paper considers three methods, i.e., the method of steepest descent, the conjugate method, and the valley searching method, and analyzes the relation between the searching method and the learning surface. It is experimentally shown based on the result of simulation that the valley searching method is better. More precisely, the parameter dependencies of the searching method on the learning curve and the searching process are examined experimentally, and it is shown that the parameters are set more easily in the valley searching method than in other methods. It is also shown that the valley searching methods is better also in terms of the convergence time and the dependency on the initial value. [ABSTRACT FROM AUTHOR]