Search

Currently, effective Intrusion-detection systems (IDS) still represent one of the important security tools. However, hybrid models based on the IDS achieve better results compared with intrusion detection based on a single algorithm. But even so, the hybrid models based on traditional algorithms still face different limitations. This work is focused on providing two main goals; firstly, analysis based on the main methods and limitations of the most-recent hybrid model-based on intrusion detection, secondly, to propose a novel hybrid IDS model called TLBO-ELM based on the Firefly algorithm and Fast Learning Network.

Many studies focusing on integrating reviews with ratings to improve recommendation performance have been quite successful. However, these works still face several shortcomings: (1) The importance of dynamically integrating review and interaction data features is typically ignored, yet treating these fusion features equally may lead to an incomplete understanding of user preferences. (2) Some forms of soft attention methods are adopted to model the local semantic information of words. As features thus captured may contain irrelevant information, the generated attention map is neither discriminatory nor detailed. In this paper, we propose a novel Adaptive Hierarchical Attention-enhanced Gated network integrating reviews for item recommendation, named AHAG. AHAG is a unified framework to capture the hidden intentions of users by adaptively incorporating reviews. Specifically, we design a gated network to dynamically fuse the extracted features and select the features that are most relevant to user preferences. To capture distinguishing fine-grained features, we introduce a hierarchical attention mechanism to learn important semantic information features and the dynamic interaction of these features. Besides, the high-order non-linear interaction of neural factorization machines is utilized to derive the rating prediction. Experiments on seven real-world datasets show that the proposed AHAG significantly outperforms state-of-the-art methods.

Conversational systems have recently attracted significant attention. Both the research community and industry believe that it will exert huge impact on human-computer interaction, and specifically, the IR/RecSys community has begun to explore Conversational Recommendation. In real-life scenarios, such systems are often urgently needed in helping users accomplishing different tasks under various situations. However, existing works still face several shortcomings: (1) Most efforts are largely confined in single task setting. (2) The conversational recommender naturally has access to the back-end data which should be fully leveraged to yield good recommendations. In this paper, we thus present a Topic-guided Conversational Recommender which is specifically designed for the multi-domain setting. It augments the sequence-to-sequence models with a neural latent topic component to better guide the response generation. To better leverage the dialogue history and the back-end data structure, we adopt a graph convolutional network to model the relationships between different recommendation candidates while also capture the match between candidates and the dialogue history. We then seamlessly combine these two parts with the idea of pointer networks. Extensive experiments show that our method achieves superior performance as compared to a wide range of baselines.

Next-generation sequencing (NGS) has been widely adopted for clinical HLA typing and advanced immunogenetics researches. Current methodologies still face challenges in resolving cis-trans ambiguity involving distant variant positions, and the turnaround time is affected by testing volume and batching. Nanopore sequencing may become a promising addition to the existing options for HLA typing. The technology delivered by the MinION sequencer of Oxford Nanopore Technologies (ONT) can record the ionic current changes during the translocation of DNA/RNA strands through transmembrane pores and translate the signals to sequence reads. It features simple and flexible library preparations, long sequencing reads, portable and affordable sequencing devices, and rapid, real-time sequencing. However, the error rate of the sequencing reads is high and remains a hurdle for its broad application. This review article will provide a brief overview of this technology and then focus on the opportunities and challenges of using nanopore sequencing for high-resolution HLA typing and immunogenetics research.

Concept-cognitive learning (CCL) is an emerging field of concerning incremental concept learning and dynamic knowledge processing in the context of dynamic environments. Although CCL has been widely researched in theory, the existing studies of CCL have one problem: the concepts obtained by CCL systems do not have generalization ability. In the meantime, the existing incremental algorithms still face some challenges that: 1) classifiers have to adapt gradually and 2) the previously acquired knowledge should be efficiently utilized. To address these problems, based on the advantage that CCL can naturally integrate new data into itself for enhancing flexibility of concept learning, we first propose a new CCL model (CCLM) to extend the classical methods of CCL, which is not only a new classifier but also good at incremental learning. Unlike the existing CCL systems, the theory of CCLM is mainly based on a formal decision context rather than a formal context. In learning concepts from dynamic environments, we show that CCLM can naturally incorporate new data into itself with a sufficient theoretical guarantee for incremental learning. For classification task and knowledge storage, our results on various data sets demonstrate that CCLM can simultaneously: 1) achieve the state-of-the-art static and dynamic classification task and 2) directly accomplish preservation of previously acquired knowledge (or concepts) under dynamic environments.

Assessing and monitoring benthic communities is increasingly important in view of global alteration of marine environments. Deep learning has proven to effectively detect marine specimen in underwater imagery but still face problems with small input datasets, unconstrained environments and class imbalance. This study evaluates a data augmentation strategy to alleviate these limitations. Through synthetically derived image compositions, the entire input dataset was greatly extended from 700 to 12700 images. Additionally, specimen numbers of brittle stars, soft corals and glass sponges are equalized resulting in a mean average precision increase of 24 %. The overall mean average precision for box detections yields 76.7 and for instance segmentation 67.7 at an intersection over union threshold of 0.5. This study shows that deep architectures such as the deployed CenterMask via ResNeXt-101 model can successfully be trained with few original images from varying underwater scenes.

The article presents the findings of main recent research and development works conducted for gas leak detection and measurement. Higher emphasis are considered for 2-D arrays sensors because of the wider coverage area they can handle in real-time, especially if they are carried within moving vehicles such as drones, helicopters, and satellites. This includes long-wave infrared (LWIR)- and medium-wave infrared (MWIR)-imaging systems. In spite of their success to accurately detect, localize, and even quantify leaks, under certain conditions, LWIR and MWIR imaging systems still face several challenges. This prevented them to continuously operate in all environmental conditions and within a reasonable long period of time. Thus, the article also presents other alternative technologies which can potentially complement or substitute LWIR and MWIR techniques. This includes photoacoustic, temperature profiling, and short-wave infrared (SWIR)-based techniques. The article can be a useful resource to help deciding which of these techniques can be used for a particular gas leak-detection application.

La implementación y aplicación de las Tecnologías de la Información y Comunicación en la educación enfrenta aún muchos desafíos; consideradas como herramientas que dan soporte al proceso enseñanza-aprendizaje presencial, ofrecen excelentes escenarios para que su efectividad y eficacia mejoren la calidad de la educación superior. El objetivo de esta investigación es analizar cómo la aplicación de estas herramientas tecnológicas dentro del aula de clases mejora en el rendimiento académico promedio de los estudiantes que cursaron la materia de Matemáticas Financieras de la carrera de Administración de Empresas de la Universidad de Guayaquil en el periodo académico 2018 – 2019 ciclos I y II, para esto se hizo un análisis descriptivo de dos muestras independientes y una aplicación de la regla de Bayes. Los resultados que arrojó esta investigación concluyen que los estudiantes que acceden y aplican herramientas tecnológicas dentro del aula tienen un mejor rendimiento académico que aquellos estudiantes que no lo aplican.

Clustering is an important unsupervised learning method in machine learning and data mining. Many existing clustering methods may still face the challenge in self-identifying clusters with varying shapes, sizes and densities. To devise a more generic clustering method that considers all the aforementioned properties of the natural clusters, we propose a novel clustering algorithm named Anchor Points based Clustering (APC). The anchor points in APC are characterized by having a relatively large distance from data points with higher densities. We take anchor points as centers to obtain intermediate clusters, which can divide the whole dataset more appropriately so as to better facilitate further grouping. In essence, based on the analysis of the identified anchor points, the relationship among the corresponding intermediate clusters can be better revealed. In short, the difference in local densities (densities within neighboring data points) of the anchor points characterizes their different properties, that is to say, all the intermediate clusters may fall into one or multiple identified levels with different densities. Finally, based on the properties of anchor points, APC spontaneously chooses the appropriate clustering strategies and reports the final clustering results. To evaluate the performances of APC, we conduct experiments on twelve two-dimensional synthetic datasets and twelve multi-dimensional real-world datasets. Moreover, we also apply APC to the Olivetti Face dataset to further assess its effectiveness in terms of face recognition. All experimental results indicate that APC outperforms four classical methods and two state-of-the-art methods in most cases. AI Singapore Ministry of Health (MOH) National Research Foundation (NRF) Accepted version This research is supported by the National Natural Science Foundation of China (61772227,61572227), the Science & Technology Development Founda- tion of Jilin Province (20180201045GX) and the Social Science Foundation of Education Department of Jilin Province (JJKH20181315SK). This research is also supported, in part, by the National Research Foundation Singapore under its AI Singapore Programme (Award Number: AISG-GC-2019-003), the Singapore Ministry of Health under its National Innovation Challenge on Active and Confident Ageing (NIC Project No. MOH/NIC/COG04/2017), and the Joint NTU-WeBank Research Centre on Fintech, Nanyang Technological University, Singapore.

Indoor pedestrian positioning has been widely used in applications such as fire rescue and indoor navigation. Compared with other indoor positioning technologies such as Wi-Fi and UWB, inertial measurement unit (IMU) based indoor positioning does not require external facilities and has lower cost. However, the major issue of IMU-based indoor positioning is that the inertial sensors exhibit errors and the errors accumulates, which affects the positioning precision. Existing IMU-based indoor positioning techniques have reduced the accumulated error, but still face several issues in positioning precision and system cost. In this paper, we propose a new IMU-based indoor pedestrian positioning technique. It adopts motion speed based adaptive error compensation and step detection based up/downstairs tracking to improve the positioning precision without using additional sensors. Compared with the existing techniques, the proposed technique achieves higher positioning precision (down to 0.11% for the closed error, 0.15% for the distance error and 0.52% for the height error) with less number of sensors (only accelerator and gyroscope) to lower the cost.

High fidelity multiphysics simulations are ubiquitous in science and engineering but still face many challenges to run efficiently in today's supercomputers. This work reports advanced technologies...

Dynamic rule mining can discover time-dependent association rules and provide more accurate descriptions about the relationship among items at different time periods and temporal granularities. However, users still face some challenges in analyzing and choosing reliable rules from the rules identified by algorithms, because of the large number of rules, the dynamic nature of rules across different time periods and granularities and the opacity of the relationship between rules and raw data. In this paper, we present our work on the development of DART, a visual analytics system for dynamic association rule mining, to help analysts gain a better understanding of rules and algorithms. DART allows users to explore rules at different time granularities (e.g., per hour, per day, per month, etc.) and with different time periods (e.g., daily, weekly, yearly, etc.), and to examine rules at multiple levels of detail, including investigating temporal patterns of a set of rules, comparing multiple rules, and evaluating a rule with raw data. Two case studies are used to show the functions and features of DART in analyzing business data and public safety data.

A major problem we still face today in this connected world, where more than 100 Quadrillion bytes of data are created every day, accelerating with the growth of the Internet of Things (IoT), needs...

Many tools have been developed for the assessment of muscle tone of impaired limbs. Despite, having the appropriate knowledge, therapists still face with challenge in the assessment due to the subjective evaluation of muscle tone during training sessions. Moreover, the training has become more-costly and time consuming since the subjects have to face the therapists over a long period of time. By deploying robot-assisted system, some of these problems could have been addressed but the aspect of proper assessment of subjects’ muscle tone levels still remain. Assessment of subjects’ muscle tones allows proper prescription of task during training session. Recent studies have established links between muscle tone and upper-limb mechanical impedance. However the development of adequate estimation algorithm for subjects’ upper-limb impedance parameters and the prediction of muscle tone level in a more objective manner is still a subject of many research works. This study proposes an appropriate strategy for the estimation of upper-limb mechanical impedance parameters as a mean for the assessment of subjects’ muscle tone levels. Both simulation and experimental results show that the upper-limb impedance parameters can be estimated to a good accuracy level, while the subjects’ muscle tone level can be consistently predicted.

In this era, use of credit has increased extensively. People are using credit cards more frequently in their daily lives. Millions of dollars were lost in forgery, however fraud detection methods have been developed to fix such types of problems. However, Nevertheless, we still face such problems through imbalance data. There is a requirement for a centralized claims database to collect views of holistic view of fraudulent characteristic behaviour. Credit card supervised learning are widely used to detect fraud based on the assumption that the pattern of fraud would depend on past transaction. This paper will give succinct understanding to avoid such problems. And will elaborate how we can use unsupervised techniques in forgery or fraud detection. In the beginning all the supervised techniques are given, types of anomalies and classification of anomaly detection techniques are given. In addition to that a project is done with the use of raw bank data and forgery is detected with unsupervised techniques.

Micro, Small, and Medium Enterprises (MSMEs) have an important role in improving the economy of small communities and the stability of the Indonesian economy. However, they still face various problems such as capital, raw materials, distribution, licensing, and others. This study developed a special treatment recommendation system to increase the independence of MSMEs by adopting a ranking-based method to produce recommendations related to problem-solving steps from MSMEs. The ranking method used Normalized Rating Frequency (NRF). This method performed an aggregation process of the ratings given by MSMEs in looking at the priority level of problem-solving based on previous experience. The result of the experiment showed that an average NDCG value was 0.7916. This explained that the quality of the ranking was good so that it was feasible to be recommended in solving problems in the form of special treatment to acquire to the independence of MSMEs.

Designing implantable bioelectronic systems that continuously monitor physiological functions and simultaneously provide personalized therapeutic solutions for patients remains a persistent challenge across many applications ranging from neural systems to bioelectronic organs. Closed-loop systems typically consist of three functional blocks, namely, sensors, signal processors and actuators. An effective system, that can provide the necessary therapeutics, tailored to individual physiological factors requires a distributed network of sensors and actuators. While significant progress has been made, closed-loop systems still face many challenges before they can truly be considered as long-term solutions for many diseases. In this review, we consider three important criteria where materials play a critical role to enable implantable closed-loop systems: Specificity, Biocompatibility and Connectivity. We look at the progress made in each of these fields with respect to a specific application and outline the challenges in creating bioelectronic technologies for the future.

With the rise of Deep Learning (DL), numerous DL-based SAR ship detectors, represented by Faster R-CNN, is constantly breaking the record of detection accuracy. However, these detectors still face huge challenges in modeling the geometric transformation of shape-changeable ships, due to their used conventional convolution kernels whose structure is fixed. Therefore, to address this problem, we propose an improved Faster R-CNN by using deformable convolution kernels for SAR ship detection. We substitute some conventional shape-changeless convolution kernels in Faster R-CNN with deformable convolution ones that can adaptively learn additional 2-D offsets of the raw convolution kernels, to better model the geometric transformation of shape-changeable ships. Finally, the experimental results on the open SAR Ship Detection Dataset (SSDD) reveal that our improved Faster R-CNN achieves a 2.02% mean Average Precision (mAP) improvement than the raw Faster R-CNN.

The potential harms and drawbacks of automated decision making has become a challenge as data science blends into our lives. In particular, fairness issues with deployed machine learning models have drawn significant attention from the research community. Despite the myriad of algorithmic fairness work in various research communities, in practice data scientists still face many roadblocks in ensuring the fairness of their machine learning models. This is primarily because there does not exist an end-to-end system that guides the users in building a fair machine learning model in a responsible way from model auditing, to model explanation, to bias mitigation. We propose a explorative model building system FairRover for responsible fair model building. FairRover guides users in (1) discovering the potential biases in the model; (2) providing explanation to the discovered biases so as to help users in understanding potential causes of the biases; and (3) mitigating the most important biases selected by the users. Because of the impossibility theorem of fairness, and the well-known trade-off between fairness and accuracy, it is generally impossible to achieve a completely fair and accurate machine learning model. Therefore, this responsible model building process is naturally performed iteratively until a satisfying trade-off is reached. Human users are involved in the loop to make various decisions guided by FairRover. We demonstrate a case study on the Adult Census dataset, which shows how FairRover guides users in iteratively building a fair income prediction model in a responsible way. We discuss the current limitations of FairRover and future work.

Extracellular vesicles (EVs) are a family of particles/vesicles present in blood and body fluids, composed of phospholipid bilayers that carry a variety of molecules that can mediate cell communication, modulating crucial cell processes such as homeostasis, induction/dampening of inflammation, and promotion of repair. Their existence, initially suspected in 1946 and confirmed in 1967, spurred a sharp increase in the number of scientific publications. Paradoxically, the increasing interest for EV content and function progressively reduced the relevance for a precise nomenclature in classifying EVs, therefore leading to a confusing scientific production. The aim of this review was to analyze the evolution of the progress in the knowledge and definition of EVs over the years, with an overview of the methodologies used for the identification of the vesicles, their cell of origin, and the detection of their cargo. The MISEV 2018 guidelines for the proper recognition nomenclature and ways to study EVs are summarized. The review finishes with a “more questions than answers” chapter, in which some of the problems we still face to fully understand the EV function and potential as a diagnostic and therapeutic tool are analyzed.

Intelligent warehouse management system (IWMS) attracts lots of interest by increasing the productivity and efficiency, minimizing the number of human workers and decreasing errors. As an important part of IWMS, warehouse portal in-out registration still face with enormous challenge. Usually, multiple UHF RFID reader antennas are installed on the portal, in-out registration can be performed by estimating the locations of RFID tags which are attached on the goods. Even though the precise localization accuracy is not required, it's still hard to distinguish the import and export directions. In this paper, we analyzed the fluctuation of received signal strength indicator (RSSI) backscattered from the RFID tags and proposed a movement direction judging method based on double-tag with an exact spacing distance. Experimental results have shown that the double-tag method can distinguish the moving direction exactly in portal environments and realize efficient in-out registration.

In the current expanding digital society, healthcare systems still face digitalization challenges due to legacy systems and the complex workflows for processing medical data. Adoption of internal Enterprise Architectures and Inter-Enterprise Architectures can provide seamless integration by allowing all participating parties to contribute to the achievement of an integrated system in accordance with the organization strategy. The authors propose a study of enabling paradigms and technologies for the design of a health 4.0 systems architecture.

Background Cereals are globally consumed as staple food, providing essential nutrients to the consumers. The emerging problems associated with cereals are mycotoxin outbreaks and adulteration in both the cereals and their products. Both classical and emerging techniques are extensively exploited to tackle such problems. However classical methods still face more limitations as compared to emerging nondestructive methods. Scope and methods High-performance liquid chromatography, gas chromatography and enzyme-linked immunosorbent assays, are classical methods employed for assessing the quality and safety parameters of cereal foods, with limitations of offline, time consuming and destructiveness. On the other hand, emerging nondestructive methodologies like hyperspectral imaging, fluorescence spectroscopy, near infrared spectroscopy and Fourier transform infrared spectroscopy have been introduced as promising techniques for the assessment of fungal contamination, quality discrimination and adulteration detection in cereals and cereal products. This review highlights the most recent applications of enlisted classical and emerging approaches, their working principles and advances in determining various safety and quality attributes of cereals and their products. Besides, challenges and future trends, their advantages and limitations are also elucidated. Key findings and conclusions Classical methods have been exploited for safety and quality measurement of cereal foods with relatively low efficiency, as compared to emerging nondestructive techniques. Classical methods suffer from disadvantages of destructiveness, thus they cannot be used for on-line monitoring, detection and evaluation. Emerging approaches are reliable with accurate, fast and non-invasive nature of investigations for the authentication of safety and quality attributes of cereal grains and their products during storage and processing. These innovative technologies can overcome the complexity, troubles, destructibility and slowness associated with classical analytical tools.

The article focuses on a few central issues of dialogic competence-in-performance which are still beyond the reach of models of Artificial Intelligence (AI). Learning machines have made an amazing step forward but still face barriers which cannot be crossed yet. Linguistics is still described at the level of Chomsky’s view of language competence. Modelling competence-in-performance requires a holistic model, such as the Mixed Game Model (Weigand 2010), which is capable of addressing the challenge of the ‘architecture of complexity’ (Simon 1962). The complex cannot be ‘the ontology of the world’ (Russell and Norwig 2016). There is no autonomous ontology, no hierarchy of concepts; it is always human beings who perceive the world. ‘Anything’, in the end, depends on the human brain.

Quantitative structure-activity relationships (QSAR) modeling is a well-known computational technique with wide applications in fields such as drug design, toxicity predictions, nanomaterials, etc. However, QSAR researchers still face certain problems to develop robust classification-based QSAR models, especially while handling response data pertaining to diverse experimental and/or theoretical conditions. In the present work, we have developed an open source standalone software "QSAR-Co" (available to download at https://sites.google.com/view/qsar-co ) to setup classification-based QSAR models that allow mining the response data coming from multiple conditions. The software comprises two modules: (1) the Model development module and (2) the Screen/Predict module. This user-friendly software provides several functionalities required for developing a robust multitasking or multitarget classification-based QSAR model using linear discriminant analysis or random forest techniques, with appropriate validation, following the principles set by the Organisation for Economic Co-operation and Development (OECD) for applying QSAR models in regulatory assessments.

Traditional recommendation methods based on matrix factorization techniques have yielded immense success because of their good scalability. However, they still face the problem of data sparsity, which may lead to a reduction in recommendation performance. As it is hard to learn good latent features in the sparse user-item rating matrix. In recent years, deep learning is very appealing in learning effective representations. Its non-linear characteristics just remedy the shortcomings of matrix factorization. In this paper, a novel method deep variational matrix factorization recommendation (DVMF) is proposed for large scale sparse dataset. DVMF is based on latent factors to predict the ratings. The latent features of the users and items are respectively obtained through a deep nonlinear structure. Based on the latent factors and combined with matrix factorization method, the paper presents algorithm optimization method of DVMF. The experiments on three real-world datasets from different domains show that DVMF is able to provide higher accuracy than recommendation algorithms based on matrix factorization or deep learning individually on large scale sparse dataset.

With Android applications (apps) becoming increasingly popular, there exist huge risks lurking in the app marketplaces as most malicious software attempt to collect users' private information without their awareness. Although these apps request users' authorization for permissions, the users can still face privacy leakage issues due to their limited knowledge in distinguishing permissions. Thus, accurate and automatic permission checking is necessary and important for users' privacy protection. According to previous studies, analyzing app descriptions is a helpful way to examine whether some permissions are required for apps. Different from those studies, we consider app permissions from a more fine-grained perspective and aim at predicting the multiple correspondent permissions to one sentence of app description. In this paper, we propose an end-to-end framework for assessing the consistency between descriptions and permissions, named Assessing Consistency based on neural Network (AC-Net). For evaluation, a new dataset involving the description-to-permission correspondences of 1415 popular Android apps was built. The experiments demonstrate that AC-Net significantly outperforms the state-of-the-art method by over 24.5% in accurately predicting permissions from descriptions.

Vehicle detection in aerial images is a challenging task and plays an important role in a wide range of applications. Traditional detection algorithms are based on sliding-window searching and shallow-learning-based features, which limits the ability to represent features and generates a lot of computational costs. Recently, with the successful application of convolutional neural network in computer vision, many state-of-the-art detectors have been developed based on deep CNNs. However, these CNN-based models still face some difficulties and challenges in vehicle detection in aerial images. Firstly, the CNN-based detection model requires extensive calculations during training and detection, and the accuracy of detection for small objects is not high. In addition, deep learning models often require a large amount of sample data to train a robust detection model, while the annotated data of aerial vehicles is limited. In this study, we propose a lightweight deep convolutional neural network detection model named LD-CNNs. The detection algorithm not only greatly reduces the computational costs of the model, but also significantly improves the accuracy of the detection. What's more, in order to cope with the problem of insufficient training samples, we develop a multi-condition constrained generative adversarial network named MC-GAN, which can effectively generate samples. The detection performance of the proposed model has been evaluated on the Munich public dataset and the collected dataset respectively. The results show that on the Munich dataset, the proposed method achieves 86.9% on mAP (mean average precision), F1-score is 0.875, and the detection time is 1.64s on Nvidia Titan XP. At present, these detection indicators have reached state-of-the-art level in vehicle detection of aerial images.

Machine reading and comprehension using differentiable reasoning models has recently been studied extensively, and memory networks have demonstrated promising performance on some reasoning tasks such as factual reasoning and basic deduction. However, as a natural language understanding model, memory networks still face challenges on the numeric representations for sentences, particularly the text representation method and the effectiveness of learned vector representations. In this paper, inspired by the convolution mechanism in the computer vision domain, a raw text representation architecture for question answering problem named convolutional end-to-end memory networks(CMemN2N) architecture is proposed. The convolutional architecture of the proposed model allows us to abstract the useful local information for reasoning to get the significant numeric sentence representation passed to the follow-up sub-tasks. Our experiments show that CMemN2N achieves better results on most of the 20 bAbI task dataset, yielding improvements for the average result compared to the state-of-the-art.

Microbial Whole-Cell Biosensors (MWCBs) have seen rapid development with the arrival of 21(st) century biological and technological capabilities. They consist of microbial species which produce, or limit the production of, a reporter protein in the presence of a target analyte. The quantifiable signal from the reporter protein can be used to determine the bioavailable levels of the target analyte in a variety of sample types at a significantly lower cost than most widely used and well-established analytical instrumentation. Furthermore, the versatile and robust nature of MWCBs shows great potential for their use in otherwise unavailable settings and environments. While MWCBs have been developed for use in biomedical, environmental, and agricultural monitoring, they still face various challenges before they can transition from the laboratory into industrialized settings like their enzyme-based counterparts. In this comprehensive and critical review, we describe the underlying working principles of MWCBs, highlight developments for their use in a variety of fields, detail challenges and current efforts to address them, and discuss exciting implementations of MWCBs helping redefine what is thought to be possible with this expeditiously evolving technology.

Until now, there is no clear definition of microtissue; it usually refers to the microtissue formed by the aggregation of seed cells under the action of cell-cell or cell-extracellular matrix (ECM). Compared with traditional cell monolayer culture, cells are cultivated into a three-dimensional microstructure in a specific way. The microstructure characteristics of microtissue are similar to natural tissues and can promote cell proliferation and differentiation. Therefore, it has a broader range of biomedical applications in tissue engineering. The traditional tissue engineering strategy is to add high-density seed cells and biomolecules on a preformed scaffold to construct a tissue engineering graft. However, due to the destruction of the ECM of the cells cultured in a monolayer during the digestion process with trypsin, the uneven distribution of the cells in the scaffold, and the damage of various adverse factors after the cells are implanted in the scaffold, this strategy is often ineffective, and the subsequent applications still face challenges. This article reviews the latest researches of a new strategy-tissue engineering microtissue strategy; discuss several traditional construction methods, structure, and function optimization; and practical application of microtissue. The review aims to provide a reference for future research on tissue engineering microtissue. Impact statement The traditional tissue engineering strategies have several disadvantages, researchers have conducted extensive research on tissue engineering microtissues in recent years, and they make significant progress. Microtissue is a kind of microtissue with three-dimensional structure, its microstructure is similar to that of natural tissue. In addition, microtissue implantation can protect cells from mechanical interference, inflammation, and other adverse factors. Furthermore, it improves the survival rate of cells and the therapeutic effect of tissue-engineered grafts. However, the practical conditions, advantages, and disadvantages of tissue engineering microtissue have not been fully elucidated. The purpose of this review is to discuss the latest research progress of microtissue and provide a reference for future research.

The assessment of three-dimensional (3D) brain cytoarchitecture at a cellular resolution remains a great challenge in the field of neuroscience and constant development of imaging techniques has become crucial, particularly when it comes to offering direct and clear obtention of data from macro to nano scales. Magnetic resonance imaging (MRI) and electron or optical microscopy, although valuable, still face some issues such as the lack of contrast and extensive sample preparation protocols. In this context, x-ray microtomography (μCT) has become a promising non-destructive tool for imaging a broad range of samples, from dense materials to soft biological specimens. It is a new supplemental method to be explored for deciphering the cytoarchitecture and connectivity of the brain. This review aims to bring together published works using x-ray μCT in neurobiology in order to discuss the achievements made so far and the future of this technique for neuroscience.

Periodic frequent pattern discovery is a non-trivial task for analysing databases to reveal the recurring shapes of patterns’ occurrences. Though significant strides have been made in their discovery for understanding large databases in decision-making, existing techniques still face a challenge of reporting a large number of periodic frequent patterns, most of which are often not useful as their periodic occurrences are either by random chance or can be inferred from the periodicities of other periodic frequent patterns. Reporting such periodic frequent patterns not only degrades the performance of existing algorithms but also could adversely affect decision-making. This study addresses these issues by proposing a novel algorithm named SRPFPM (Self-Reliant Periodic Frequent Pattern Miner) for mining and reporting the set of self-reliant periodic frequent patterns as those whose periodic occurrences have inherent item relationships and cannot be inferred from other periodic frequent patterns. Experimental analysis on benchmark datasets show that SRPFPM is efficient and effectively prunes periodic frequent patterns that are periodic due to random chance as well as those whose periodicities can be inferred from other periodic frequent patterns.

The past 30 years have seen increased adoption of learning management systems (LMSs) in education. Several studies have investigated the usability of LMSs for students. However, very few studies have assessed the usability from the instructors’ perspective. Usability issues can pose challenges for instructors who use LMSs to create, manage and deliver courses. These challenges require instructors to spend extra time and energy on tackling the challenges rather on teaching-related tasks, which will have negative impacts on learners’ experiences and learning outcomes. This paper aims to identify usability challenges in LMSs for instructors. We used Canvas as an example and conducted user testing with 35 university instructors in computer science and engineering disciplines. Pre- and post-interviews were transcribed and analyzed together with the observation data during their use of Canvas to carry out tasks. The results show that, although Canvas has made continuous efforts to improve its usability, instructors still face some usability challenges. Instructors are a diverse user group for LMSs. Further research should consider recruiting participants from other disciplines and investigating other LMSs to identify possibilities for improving general usability of digital tools for instructors.

Cycling is on the rise as a relevant alternative to car-based mobility and even though there are mobile applications specifically designed for cyclists to support this development, many still face unresolved challenges in terms of safe user interaction with complex data while riding. We present the design, development, and evaluation of VeloCity - an application for reporting traffic incidents and structures relevant to cyclists. In a case study, we compared its’ three input methods (touch, in-app speech recognition, the voice assistant of the operating system) to evaluate which attributes make for safe interaction while cycling. We found that participants prefer to use the voice assistant over the other modalities as it was the least distracting due to its hands- and eyes-free interaction design. Furthermore, they chose short commands over conversational phrases. Based on our results, we present five guidelines for designing voice user interfaces for cyclists and argue for moving away from touch-based interfaces in this domain, which still make up most of the applied interaction techniques today.

Relational triple extraction is critical to understanding massive text corpora and constructing large-scale knowledge graph, which has attracted increasing research interest. However, existing studies still face some challenging issues, including information loss, error propagation and ignoring the interaction between entity and relation. To intuitively explore the above issues and address them, in this paper, we provide a revealing insight into relational triple extraction from a stereoscopic perspective, which rationalizes the occurrence of these issues and exposes the shortcomings of existing methods. Further, a novel model is proposed for relational triple extraction, which maps relational triples to a three-dimension (3-D) space and leverages three decoders to extract them, aimed at simultaneously handling the above issues. A series of experiments are conducted on five public datasets, demonstrating that the proposed model outperforms the recent advanced baselines.

Although the trajectory data mining has made remarkable achievements in the development of smart urban services, we still face a few of unavoidable open issues which need to be further explored. In this chapter, we will discuss some open issues from the perspectives of data, model, and application. Afterwards, we chart several promising future directions of the GPS trajectory data mining and smart urban service developing, with hoped-for deepening and broadening the intelligence of smart city. Finally, we make conclusions of this monograph.

With the launch of 3GPP fifth-generation (5G) commercial cellular networks around the world, the research community has started focusing on the design of the sixth-generation (6G) system. One of the considerations is the use of Terahertz communications that aims to provide 1 Tbps (terabits per second) and air latency less than 100 μs. Further, 6G networks are expected to provide for more stringent Quality of Service (QoS) and mobility requirements. While addition to innovations at the physical layer and radio technologies can achieve these goals to a great extent, the end-to-end applications would still face challenges to fully utilize the network capacity due to limitations of the current transport layer protocols. In this chapter, we explore the challenges in the design of next-generation transport layer protocols (NGTP) in 6G Terahertz communication-based networks. Some of the challenges are due to user mobility, high-speed and high-bitrate communications, and other issues. The impact of these issues and potential approaches to mitigate these challenges are also discussed.

Conventional prostheses are incapable of reproducing full functionality of biological limbs. Continuous advancements in robotics and materials science have led to the development of soft and anthropomorphic prosthetic hands and legs. Mimicking the compliance and structure of biological limbs provides dexterity to upper limb prosthetic users, and natural gait to lower limb prosthetic users. Although soft and anthropomorphic prosthetic technology has reached a certain maturity level, technologies for restoring somatosensation still face significant challenges. Providing somatosensory feedback can improve the quality of life of amputees by augmenting the functionality of prostheses. Advanced prosthetic sensors obtain various sensory information, while ensuring compliant interaction with the environment. The development of electronic skins that combine multiple sensors and mimic functionalities of biological skin is possible with the recent advancements in materials technology. This chapter reviews soft and anthropomorphic upper and lower limb prostheses, prosthetic sensors, electronic skins, and applications of prosthetic interfaces.

Internet of Things (IoT) devices are spread in different areas such as e-tracking, e-commerce, e-home, and e-health, etc. Thus, during the last ten years, the internet of things technology (IoT) has been a research focus. Both privacy and security are the key concerns for the applications of IoT, and still face a huge number of challenges. There are many elements used to run the IoT technology which include hardware and software such as sensors, GPS, cameras, applications, and so forth. In this paper, we have analyzed and explain the technology of IoT along with its elements, security features, security issues, and threats that attached to each layer of IoT to guide the consideration of researchers into solve and understand the most serious problems in IoT environment.

Locomotion is essential for a person’s ability to function in society. When an individual has a condition that limits locomotion, such as a lower limb amputation, the performance of a prosthetic often determines the quality of life an individual regains. In recent years, powered prosthetic devices have shown nearly identical replication for human leg motion on non-compliant terrains. However, they still face numerous functional deficits such as increased metabolic cost and instability for walking on surfaces of varying compliance and complexity. This paper proposes joint angles of the biological leg are uniquely altered by surface compliance regardless of a subject’s individual walking pattern. These differences are then displayed and quantified as a way to better characterize able-bodied walking compensation typical with three common terrains: sand, grass and gravel. This study also collects data outdoors using IMU sensors and is not limited by lab setup and conditions. These results are important since better understanding of joint angle kinematics on varying terrains could enable the formulation of advanced controllers for current prosthetic devices allowing them to anticipate surface changes and adapt accordingly.

Despite the current diversity and inclusion initiatives in the academic community, researchers with a non-native command of English still face significant obstacles when writing papers in English. This paper presents the Langsmith editor, which assists inexperienced, non-native researchers to write English papers, especially in the natural language processing (NLP) field. Our system can suggest fluent, academic-style sentences to writers based on their rough, incomplete phrases or sentences. The system also encourages interaction between human writers and the computerized revision system. The experimental results demonstrated that Langsmith helps non-native English-speaker students write papers in English. The system is available at https://emnlp-demo.editor. langsmith.co.jp/., Accepted at EMNLP 2020 (system demonstrations)

State-of-the-art semantic segmentation methods require sufficient labeled data to achieve good results and hardly work on unseen classes without fine-tuning. Few-shot segmentation is thus proposed to tackle this problem by learning a model that quickly adapts to new classes with a few labeled support samples. Theses frameworks still face the challenge of generalization ability reduction on unseen classes due to inappropriate use of high-level semantic information of training classes and spatial inconsistency between query and support targets. To alleviate these issues, we propose the Prior Guided Feature Enrichment Network (PFENet). It consists of novel designs of (1) a training-free prior mask generation method that not only retains generalization power but also improves model performance and (2) Feature Enrichment Module (FEM) that overcomes spatial inconsistency by adaptively enriching query features with support features and prior masks. Extensive experiments on PASCAL-5$^i$ and COCO prove that the proposed prior generation method and FEM both improve the baseline method significantly. Our PFENet also outperforms state-of-the-art methods by a large margin without efficiency loss. It is surprising that our model even generalizes to cases without labeled support samples. Our code is available at https://github.com/Jia-Research-Lab/PFENet/., Comment: 16 pages. To appear in TPAMI

In Brazil, only 15% of deaf people are fluent in Brazilian Portuguese. Although Brazilian law says that public spaces should provide access in Libras, deaf still face lots of issues accessing customer services. Also, there is no tool able to allow them to talk in Libras via customer service communication channels, such as chats. Thus, this work aims to help deaf to access the customer services of a big computer manufacturer by providing a way for them to communicate in Libras while the call center attendant receives a translation in Brazilian Portuguese. The method to translate Libras to Portuguese includes the use of a CNN. The results in recognizing and classifying signs (captured from video) reached over 90% of the accuracy in the training phase.

The automatic dialogue system is an intelligent system built by combining various artificial intelligence technologies. In recent years, with the introduction of multi-turn dialogue generation systems, semantic relevance and topic consistency between different dialogue turns have become important evaluation criteria for the success of the model. However, these problems have not been resolved and still face many challenges. In this paper, we propose a sequence to sequence (seq2seq) model based on multi-encoder structure and theme-oriented decoder, and these innovations enable the proposed model to obtain semantic correlation between different turns of conversation and maintain the consistency of topics. Meanwhile, aiming at the disadvantages of the basic seq2seq model, BiLSTM cells, attention mechanism and beam search algorithm are adopted to solve the problem of long-distance dependence and obtain richer semantic information. Experiments show that the proposed model can generate coherent, appropriate and diverse replies on multi-turn dialogue datasets.

International audience; Digital Twins have emerged since the beginning of this millennium to better support the management of systems based on (real-time) data collected in different parts of the operating systems. Digital Twins have been successfully used in many application domains, and thus, are considered as an important aspect of Model-Based Systems Engineering (MBSE). However, their development , maintenance, and evolution still face major challenges, in particular: (i) the management of heterogeneous models from different disciplines, (ii) the bi-directional synchronization of digital twins and the actual systems, and (iii) the support for collaborative development throughout the complete life-cycle. In the last decades, the Model-Driven Engineering (MDE) community has investigated these challenges in the context of software systems. Now the question arises, which results may be applicable for digital twin engineering as well. In this paper, we identify various MDE techniques and technologies which may contribute to tackle the three mentioned digital twin challenges as well as outline a set of open MDE research challenges that need to be addressed in order to move towards a digital twin engineering discipline.

Many enterprises run Network Function Virtualization (NFV) services on public clouds to relieve management burdens and reduce costs. However, NFV operators still face the burden of choosing the right types of virtual machines (VMs) for various network functions (NFs), as well as the cost of renting VMs at a granularity of months or years while many VMs remain idle during valley hours. A recent computing model named serverless computing automatically executes user-defined functions on requests arrival, and charges users based on the number of processed requests. For NFV operators, serverless computing has the potential of completely relieving NF management burden and significantly reducing costs. Nevertheless, naively exploring existing serverless platforms for NFV introduces significant performance overheads in three aspects, including high remote state access latency, long NF launching time, and high packet delivery latency between NFs. To address these problems, we propose Serpens, a high-performance serverless platform for NFV. Firstly, Serpens designs a novel state management mechanism to support local state access. Secondly, Serpens proposes an efficient NF execution model to provide fast NF launching and avoid extra packet delivery. We have implemented a prototype of Serpens. Evaluation results demonstrate that Serpens could significantly improve performance for NFs and service function chains (SFCs) comparing to existing serverless platforms.

Mobile Crowdsensing (MCS) is a new paradigm that recruits users to cooperatively perform a sensing task. When recruiting users, existing works mainly focus on selecting a group of users with the best objective ability, e.g., the user’s probability or frequency of covering the task locations. However, we argue that, for the cooperative MCS task, the completion effect depends not only on the user’s objective ability, but also on their subjective collaboration likelihood with each other. In other words, in each single round, we prefer to recruit users with not only a strong objective ability but also good collaboration likelihood. Moreover, even though we could find a well-behaved group of users in a single round, in the multi-round scenario without enough prior knowledge, we still face the problem of recruiting previously well-behaved user groups (exploitation) or recruiting unknown user groups (exploration). To address these problems, in this paper, we first convert the single-round user recruitment problem into the min-cut problem and propose a graph theory based algorithm to find the optimal group of users. Furthermore, in the multi-round scenario, to balance the trade-off between exploration and exploitation, we propose the multi-round User Recruitment strategy based on the combinatorial Multi-armed Bandit model (URMB) and prove that it can achieve a tight regret bound. Finally, extensive experiments on three real-world datasets validate that the users recruited by URMB result in a better task completion effect than the state-of-the-art strategy.

Over the past few decades, the demand for the capacity and reliability of optical networks has continued to grow. In the meantime, optical networks with larger knowledge scales have become sources of numerous heterogeneous data. In order to handle these new challenges, many issues need to be resolved, among which the low-margin optical networks design, power optimization, routing and wavelength assignment (RWA), failure management are quite important. However, the use of traditional algorithms in the above four applications shows some shortcomings. Fortunately, artificial intelligence (AI), especially machine learning (ML), is regarded as one of the most promising methods to overcome these shortcomings. In this study, we review the applications of ML methods in solving these four issues. Although many ML-based researches have emerged, the applications of ML techniques in optical networks still face challenges. Therefore, we also discuss some possible future directions of investigating ML-based approaches in optical networks.

Large-scale omics experiments have become standard in biological studies, leading to a deluge of data. However, researchers still face the challenge of connecting changes in the omics data to changes in cell functions, due to the complex interdependencies between genes, proteins and metabolites. Here we present a novel framework that begins to overcome this problem by allowing users to infer how metabolic functions change, based on omics data. To enable this, we curated and standardized lists of metabolic tasks that mammalian cells can accomplish. We then used genome-scale metabolic networks to define gene modules responsible for each specific metabolic task. We further developed a framework to overlay omics data on these modules to predict pathway usage for each metabolic task. The proposed approach allows one to directly predict how changes in omics experiments change cell or tissue function. We further demonstrated how this new approach can be used to leverage the metabolic functions of biological entities from the single cell to their organization in tissues and organs using multiple transcriptomic datasets (human and mouse). Finally, we created a web-based CellFie module that has been integrated into the list of tools available in GenePattern (www.genepattern.org) to enable adoption of the approach.