Search

Exam paper generation is an indispensable part of teaching. Existing methods focus on the use of question extraction algorithms with labels for each question provided. Obviously, manual labeling is inefficient and cannot avoid label bias. Furthermore, the quality of the exam papers generated by the existing methods is not guaranteed. To address these problems, we propose a novel approach to generating exam papers based on prediction of exam performance. As such, we update the quality of the initially generated questions one by using dynamic programming, as well as in batches by using genetic algorithms. We performed the prediction task by using Deep Knowledge Tracing. Our approach considered the skill weight, difficulty, and distribution of exam scores. By comparisons, experimental results indicate that our approach performed better than the two baselines. Furthermore, it can generate exam papers with adaptive difficulties closely to the expected levels, and the related student exam scores will be guaranteed to be relatively reasonable distribution. In addition, our approach was evaluated in a real learning scenarios and shows advantages.

In order to satisfy the demands for both safety and global warming reduction, a high-strength seismic reinforced steel bar is required in the structural steel market. Recent developments in computational thermodynamics and related application software have made it possible to design a suitable material as well as support engineers of steel manufacturing companies in the production of the designed material with minimum benchmarks in practical operations. This paper reports our recent success in developing grade 600 MPa reinforced steel bars for seismic safety in South Korea. First, conventional alloy design based on CALPHAD-type computational thermodynamics was carried out. For this purpose, a typical alloy system of Fe-0.30C-0.23Si-1.37Mn-0.14V-0.22Cu (in wt%) was selected, and thermodynamic and kinetic calculations were carried out using MatCalc and JMatPro software. Second, in order to reduce V content in the steel for economic reasons, a cooling process designed using finite element (FE) simulation based on the thermodynamic database was performed. For this application, Fe-0.34C-0.22Si-1.34Mn-0.04 V (in wt%) alloy was chosen, and the FE software ABAQUS was applied for modeling the TempCore process. The mechanical properties of the steel products with a diameter of 32 mm produced based on the simulated results satisfy the required properties for grade 600 MPa seismic reinforced steel bars.

The top 1000 biomedical papers by number of citations are classified by method, type of method and non-methods by examination of citation contexts. Supervised machine learning is applied to the context data for a training sample of papers which is then used to classify the full list, revealing that words indicating utility are most important for the classification of methods. Further word analysis is carried out using corpus linguistics to uncover context words that characterize non-methods. Hedging words are found to play an important role for non-methods, and several are selected for further analysis with logistic regression. Other variables in the regression are a consensus variable based on the similarity of contexts for a paper and another variable based on whether citations come from “methods” sections of citing papers. Accuracy of predictions from logistic regression is comparable to machine learning. The results are interpreted in terms of the perceived certainty or uncertainty of the underlying knowledge, that is, methods and their outputs have higher certainty, and non-methods higher uncertainty. Evidence is found that hedging is inversely related to citation frequency. Implications of this work for the study of the development of science and the role of methods and tools in biomedical research are discussed.

This study focuses on a recently introduced type of indicator measuring disruptiveness in science. Disruptive research diverges from current lines of research by opening up new lines. In the current study, we included the initially proposed indicator of this new type (Funk & Owen-Smith, 2017; Wu, Wang, & Evans, 2019) and several variants with DI1: DI5, DI1n, DI5n, and DEP. Since indicators should measure what they propose to measure, we investigated the convergent validity of the indicators. We used a list of milestone papers, selected and published by editors of Physical Review Letters, and investigated whether this human (experts)-based list is related to values of the several disruption indicators variants and – if so – which variants show the highest correlation with expert judgements. We used bivariate statistics, multiple regression models, and (coarsened) exact matching (CEM) to investigate the convergent validity of the indicators. The results show that the indicators correlate differently with the milestone paper assignments by the editors. It is not the initially proposed disruption index that performed best (DI1), but the variant DI5 which has been introduced by Bornmann, Devarakonda, Tekles, and Chacko (2020a). In the CEM analysis of this study, the DEP variant – introduced by Bu, Waltman, and Huang (in press) – also showed favorable results.

Paper cooperation network embodies expert topic similarity in an extent, thus, a novel method is proposed for expert topic analysis and extraction by combining paper cooperation network and topic model. In the method, we extract each paper’ author information and construct an expert cooperation network. At the same time, by means of LDA model, a probabilistic topic model is also built to analyze papers’ latent topics. Then, by making full use of the feature that adjacent nodes in the expert cooperation network share similar themes distribution, we makes a constraint on expert topic distribution in Gibbs sampling process of solving the probabilistic topic model. Experimental results on NIPS dataset show that the proposed method can effectively extract expert topics, and the expert paper cooperation network plays a very good supporting role on the extracting task.

This research is concerned with the detection of similar academic papers. Given a tested paper from a given corpus of 10,099 peer-reviewed scientific papers, a two-stage process was activated. During the first stage, most of the papers were filtered out using a fast filter method. In the second stage, in order to detect similar papers we applied 23 heuristic variants derived from 3 novel prototype methods using various parameter settings. The three novel prototype methods are: CT-TR Constant Number of randomized T fingerprints, compared to each one-third of R (first/middle/last) fingerprints, CT-AR: Constant Number of randomized T fingerprints, compared to all R fingerprints, and CDT-AR: Constant Number of divided randomized T fingerprints compared, to all R fingerprints. Results achieved by the new methods are superior to those of previous heuristic methods, which were approximations of the Full Fingerprint (FF) method, currently considered the best heuristic method. The order of this new methods' run-time, (n), is far more efficient than the order of the FF method run-time, (n2) (after removing short documents from the corpus).

Although software has helped researchers conduct research, little is known of the impact of software on science. To fill this gap, this article proposes an improved bootstrapping method to extract software entities from full-text papers and assess their impact on science. Evaluation results show that the proposed entity extraction system outperforms three baseline methods on extracting software entities from full-text papers. The proposed method is then used to learn software entities from all papers published in PLoS ONE in 2014. More than 2000 unique software entities are obtained which accounted for more than 20,000 mentions and more than 7000 citations. The paper finds that software is commonly used in the scientific community along with a substantial uncitedness.

Section identification is an important task for library science, especially knowledge management. Identifying the sections of a paper would help filter noise in entity and relation extraction. In this research, we studied the paper section identification problem in the context of Chinese medical literature analysis, where the subjects, methods, and results are more valuable from a physician's perspective. Based on previous studies on English literature section identification, we experiment with the effective features to use with classic machine learning algorithms to tackle the problem. It is found that Conditional Random Fields, which consider sentence interdependency, is more effective in combining different feature sets, such as bag-of-words, part-of-speech, and headings, for Chinese literature section identification. Moreover, we find that classic machine learning algorithms are more effective than generic deep learning models for this problem. Based on these observations, we design a novel deep learning model, the Structural Bidirectional Long Short-Term Memory (SLSTM) model, which models word and sentence interdependency together with the contextual information. Experiments on a human-curated asthma literature dataset show that our approach outperforms the traditional machine learning methods and other deep learning methods and achieves close to 90% precision and recall in the task. The model shows good potential for use in other text mining tasks. The research has significant methodological and practical implications.

Previously, plant-wide disturbance analysis has looked into the propagation of faults through an industrial production process by investigating process measurements. However, the extent of the analysis has mostly been limited to a section of a plant. In this work, we propose a top-down approach which investigates measurements of the complete plant and identifies a section where the disturbance originates. Root cause analysis is carried out thereafter to pinpoint the faulty asset. The proposed approach has three novel elements: Using key performance indicators (KPI) as reference and starting point of the analysis, restricting measurements to a measurement type (e.g. flow) thus focusing on a section and applying the novel method of contribution plots of spectral PCA T 2 statistic to find the contribution of each measurement towards the disturbance observed in the KPI. The approach is described and carried out on a paper machine where a quality KPI showed an established oscillation.

Every product that exists, ranging from a toothbrush to a car, has first been conceived as a mental concept. Due to its efficacy in rapidly externalizing concepts, paper-based sketching is still extensively used by practising designers to gradually develop the three-dimensional (3D) geometric form of a concept. It is a common practice that form concepts are sketched on paper prior to generating 3D virtual models in commercial Computer-Aided Design (CAD) systems. However, the user-interface of such systems does not support automatic generation of 3D models from sketches. Furthermore, the inherent characteristics of form sketching (e.g. idiosyncrasy) pose a challenge to computer-based understanding of the form concept semantics expressed on paper. To address these issues, this paper is therefore concerned with the development of a visual language that is prescribed and to be used by product designers to annotate paper-based sketches such that the form geometry semantics can be formally represented; parsing the annotated sketch allows for the automatic generation of 3D virtual models in CAD. Inspired by re-usable 3D CAD modelling functions and the related environmental constraints and requirements, a prescribed sketching language, PSL, has been developed to annotate paper-based form sketches. The framework architecture which parses the annotated sketch and subsequently extracts the form concept semantics is described. Based on this framework, a prototype computer tool has been implemented and evaluated. Evaluation results provide a degree of evidence, first on the suitability of PSL in representing the semantics of a range of forms, and secondly on the designers? acceptance of taking up this annotated sketching approach in practice. The key contribution is a visual language to formally represent form geometry semantics on paper.Parsing the language allows for the automatic generation of 3D virtual models.A proof-of-concept prototype tool was implemented.The language is capable to roughly model forms with linear topological ordering.Evaluation results show that practising designers would use the language.

In a very recently presented paper, Turky and Abdullah 5 proposed a novel multi-population harmony search with external archive (MHSA-ExtArchive) for dynamic optimization problems. In the experimental results, the authors claimed that their approach could outperform several state-of-the-art algorithms. They also showed the superiority of their method by means of numerical experiments on Moving Peaks Benchmark (MPB). Despite the interesting idea of applying multi-population scheme on harmony search and using a new type of external archive for dealing with dynamic problems, we believe that there are two very important shortcomings in the result analysis, which we point out in this short note. The main motivation of the present note is to contribute toward preventing the same mistakes from happening by the other researchers.

Archaeologists can now collect an inordinate amount of 3D data. But are these 3D data sustainable? Are they being managed to make them accessible? The MayaArch3D Project researched and addressed these questions by applying best practices to build four prototype tools to store, manage, visualize, and analyze multi-resolution, geo-referenced 3D models in a web-based environment. While the technical aspects of these tools have been published, this position paper addresses a catch 22 that we, as archaeologists, encounter in the field of 3D archaeology – one that formed the initial impetus for the MayaArch3D Project: that is, while the quantity of 3D archaeological data is increasing, these data are not usually accessible. By researching and addressing 3D data integration and accessibility, we learned many lessons that group around four main issues: sensitivity/security, web-based dissemination, conveying uncertainty, and data storage/reuse/peer review. These are significant current challenges to making 3D archaeological data sustainable.

As the volume of medicinal information stored electronically grows, the necessity to boost the security increases too. Health care systems are now very common in the world. A healthcare system must maintain privacy while preventing patient information from being transmitted. Therefore, securing such systems is crucial for gaining public confidence in the use of network-based healthcare systems. Information hiding is particularly used for security applications to protect the secret message from an unauthorized person. Employing information hiding in healthcare systems, deals with storing, sharing, and managing medical data in a reliable and secure manner. This paper reviews the algorithms used to provide security in healthcare systems. Furthermore, these algorithms are categorized in this paper based on the application and the way of providing security. Finally, the challenges in implementing the security of healthcare systems are described.

When producing complex and highly customisable products in low volumes (or in ‘batch sizes of one’), automation of production systems is critical for competitiveness and profitability in high labour-cost economies. To facilitate batch-size-of-one production, ‘topology generation’, ‘realisability’, and ‘control’ algorithms have been developed as part of the Evolvable Assembly Systems (EAS) project. The topology generation algorithm computes all the possible sequences of parallel activities that assembly resources can perform on parts and is run offline whenever the layout of the production facility changes, whereas realisability checking and controller generation are performed at run-time to check whether a production facility with a given set of assembly resources can assemble a desired product, and how the product should be assembled, e.g., which resources to use, and when. Generated controllers are output in Business to Manufacturing Markup Language (B2MML). Taken together, the algorithms thus represent a step toward a complete path from the formal specification of an assembly system and the products to be assembled, to the automated synthesis of executable process plans. This paper presents each algorithm in sufficient detail to allow their reimplementation by other researchers. Topology generation is the most expensive step in the approach. A preliminary experimental evaluation of the scalability of topology generation is presented, which suggests that, for small to medium sized production facilities, the time required for recomputing the topology is sufficiently small not to preclude frequent factory transformations, e.g., the addition of new resources.

Today there is a rapid development taking place in phenotyping of plants using non-destructive image based machine vision techniques. Machine vision based plant phenotyping ranges from single plant trait estimation to broad assessment of crop canopy for thousands of plants in the field. Plant phenotyping systems either use single imaging method or integrative approach signifying simultaneous use of some of the imaging techniques like visible red, green and blue (RGB) imaging, thermal imaging, chlorophyll fluorescence imaging (CFIM), hyperspectral imaging, 3-dimensional (3-D) imaging or high resolution volumetric imaging. This paper provides an overview of imaging techniques and their applications in the field of plant phenotyping. This paper presents a comprehensive survey on recent machine vision methods for plant trait estimation and classification. In this paper, information about publicly available datasets is provided for uniform comparison among the state-of-the-art phenotyping methods. This paper also presents future research directions related to the use of deep learning based machine vision algorithms for structural (2-D and 3-D), physiological and temporal trait estimation, and classification studies in plants.

Image recognition based on deep learning methods has gained remarkable achievements by feeding with abundant training data. Unfortunately, collecting a tremendous amount of annotated images is time-consuming and expensive, especially in grocery product recognition tasks. It is challenging to recognise grocery products accurately when the deep learning model is trained with insufficient data. This paper proposes multi-angle Generative Adversarial Networks (MAGAN), which can generate realistic training images with different angles for data augmentation. Mutual information is employed in the novel GAN to achieve the learning of angles in an unsupervised manner. This paper aims to create training images containing grocery products from different angles, thus improving grocery product recognition accuracy. We first enlarge the fruit dataset by using MAGAN and the state-of-the-art GAN variants. Then, we compare the top-1 accuracy results from CNN classifiers trained with different data augmentation methods. Finally, our experiments demonstrate that the MAGAN exceeds the existing GANs for grocery product recognition tasks, obtaining a significant increase in the accuracy.

Video-based person re-identification plays a critical role in intelligent video surveillance by learning temporal correlations from consecutive video frames. Most existing methods aim to solve the challenging variations of pose, occlusion, backgrounds and so on by using attention mechanism. They almost all draw attention to the occlusion and learn occlusion-invariant video representations by abandoning the occluded area or frames, while the other areas in these frames contain sufficient spatial information and temporal cues. To overcome these drawbacks, this paper proposes a comprehensive attention mechanism covering what, where, and when to pay attention in the discriminative spatial-temporal feature learning, namely What-Where-When Attention Network (W3AN). Concretely, W3AN designs a spatial attention module to focus on pedestrian identity and obvious attributes by the importance estimating layer (What and Where), and a temporal attention module to calculate the frame-level importance (when), which is embedded into a graph attention network to exploit temporal attention features rather than computing weighted average feature for video frames like existing methods. Moreover, the experiments on three widely-recognized datasets demonstrate the effectiveness of our proposed W3AN model and the discussion of major modules elaborates the contributions of this paper.

The research on intelligent fault diagnosis has yielded remarkable achievements based on artificial intelligence-related technologies. In engineering scenarios, machines usually work in a normal condition, which means limited fault data can be collected. Intelligent fault diagnosis with small & imbalanced data (S&I-IFD), which refers to build intelligent diagnosis models using limited machine faulty samples to achieve accurate fault identification, has been attracting the attention of researchers. Nowadays, the research on S&I-IFD has achieved fruitful results, but a review of the latest achievements is still lacking, and the future research directions are not clear enough. To address this, we review the research results on S&I-IFD and provides some future perspectives in this paper. The existing research results are divided into three categories: the data augmentation-based, the feature learning-based, and the classifier design-based. Data augmentation-based strategy improves the performance of diagnosis models by augmenting training data. Feature learning-based strategy identifies faults accurately by extracting features from small & imbalanced data. Classifier design-based strategy achieves high diagnosis accuracy by constructing classifiers suitable for small & imbalanced data. Finally, this paper points out the research challenges faced by S&I-IFD and provides some directions that may bring breakthroughs, including meta-learning and zero-shot learning.

In this survey, a systematic literature review of the state-of-the-art on emotion expression recognition from facial images is presented. The paper has as main objective arise the most commonly used strategies employed to interpret and recognize facial emotion expressions, published over the past few years. For this purpose, a total of 51 papers were analyzed over the literature totaling 94 distinct methods, collected from well-established scientific databases (ACM Digital Library, IEEE Xplore, Science Direct and Scopus), whose works were categorized according to its main construction concept. From the analyzed works, it was possible to categorize them into two main trends: classical and those approaches specifically designed by the use of neural networks . The obtained statistical analysis demonstrated a marginally better recognition precision for the classical approaches when faced to neural networks counterpart, but with a reduced capacity of generalization . Additionally, the present study verified the most popular datasets for facial expression and emotion recognition showing the pros and cons each and, thereby, demonstrating a real demand for reliable data-sources regarding artificial and natural experimental environments.

Machine reading comprehension (MRC) is a challenging task in natural language processing (NLP), which requires machine to determine the corresponding answer to a given passage and question. Whereas, there always exist unanswerable questions in the real world, which poses a new challenge to MRC tasks. Abundant research work has been carried out on the verification mechanism for the answerability of passage-question pairs. However, these researches only focus on its design and implementation, which has limitations in real-world scenarios. Thus, the method proposed in this paper not only verifies the answerability, but also validates and adjusts the predicted answer to obtain an elaborate answer span. Using powerful pre-trained model as encoder block, this paper explores a more comprehensive verification mechanism. Similar to how humans read passages and give answers, we propose a three-stage mechanism called ”Verification for an Elaborate Span” (V4ES): 1) sketchy reading that the model briefly browses the overall information of the passage and question, and then generates an initial answer; 2) intensive reading that it reads the passage and question again, judges the answerability of the question and gives an answer at this stage; 3) verification that it verifies these two answers produced at the previous two stages, and then gives the final prediction. Moreover, the proposed model is evaluated on two MRC challenge datasets: SQuAD2.0 and CMRC2018, and the experiment results show that our model has achieved great improvement compared with the ALBERT and BERT baselines. In conclusion, our proposed verification mechanism has demonstrated its effectiveness through a series of experiments and analysis.

This paper focuses on two related sub-tasks of aspect-based sentiment analysis, namely aspect-term extraction and aspect sentiment classification. The former aims to extract aspect-terms from given sentences and the latter aims to identify the sentiment polarity expressed on the extracted terms. Considering the practical application, researchers use more joint methods rather than pipeline methods. However, existing joint methods cannot model the interaction between aspect-terms and the sentence they belong to, or consider the relevance among the sentiments of different aspect-terms. In this paper, a novel end-to-end generative model based on encoder-decoder, namely Joint Term-Sentiment Generator (JTSG), is presented to generate all aspect term-polarity pairs. Specifically, a pre-trained model based encoder is used to encode the sentences, and specially, the decoder generates the start and end position to determine an aspect-term, rather than generate aspect-terms themselves. This new generative method contributes to avoid generating incomplete aspect-terms. Experimental results demonstrate that the proposed approach yields competitive performance on three benchmark datasets.

Many methods achieve the visual object tracking task with deep learning technologies. As the deep features of different levels contain various semantic information and functions, this paper presents a multi-hierarchy feature aggregation approach to tackle the specific issues in the tracking task, which consists of two aspects. On one hand, this paper integrates the features captured by the offline and online classifiers at the score level, which constructs complementary roles of these classifiers to enhance the stability of classification. Besides, the proposed offline classifier is continuously optimized with different levels of features to reinforce classification constraints. On the other hand, we design a butterfly attention module to promote the capacity of multi-hierarchy feature aggregation in the regression network, which aims to fuse and strengthen the multi-scale features by attending to their spatial information. It can capture more spatial contexts by utilizing the self-attention mechanism during the fusion procedure, and preserve the hierarchy of the features during the strengthening process. Extensive experiments on four public datasets, i.e., VOT2018, OTB100, NFS and LaSOT datasets, demonstrate the effectiveness of the proposed methods.

Context and rationale Intrusion Detection, the ability to detect malware and other attacks, is a crucial aspect to ensure cybersecurity. So is the ability to identify this myriad of attacks. Objective Artificial Neural Networks (as well as other machine learning bio-inspired approaches) are an established and proven method of accurate classification. ANNs are extremely versatile – a wide range of setups can achieve significantly different classification results. The main objective and contribution of this paper is the evaluation of the way the hyperparameters can influence the final classification result. Method and results In this paper, a wide range of ANN setups is put to comparison. We have performed our experiments on two benchmark datasets, namely NSL-KDD and CICIDS2017. Conclusions The most effective arrangement achieves the multi-class classification accuracy of 99.909% on an established benchmark dataset.

Single-object tracking is regarded as a challenging task in computer vision, especially in complex spatio-temporal contexts. The changes in the environment and object deformation make it difficult to track. In the last 10 years, the application of correlation filters and deep learning enhances the performance of trackers to a large extent. This paper summarizes single-object tracking algorithms based on correlation filters and deep learning. Firstly, we explain the definition of single-object tracking and analyze the components of general object tracking algorithms. Secondly, the single-object tracking algorithms proposed in the past decade are summarized according to different categories. Finally, this paper summarizes the achievements and problems of existing algorithms by analyzing experimental results and discusses the development trends.

Active fire detection in satellite imagery is of critical importance to the management of environmental conservation policies, supporting decision-making and law enforcement. This is a well established field, with many techniques being proposed over the years, usually based on pixel or region-level comparisons involving sensor-specific thresholds and neighborhood statistics. In this paper, we address the problem of active fire detection using deep learning techniques. In recent years, deep learning techniques have been enjoying an enormous success in many fields, but their use for active fire detection is relatively new, with open questions and demand for datasets and architectures for evaluation. This paper addresses these issues by introducing a new large-scale dataset for active fire detection, with over 150,000 image patches (more than 200 GB of data) extracted from Landsat-8 images captured around the world in August and September 2020, containing wildfires in several locations. The dataset was split in two parts, and contains 10-band spectral images with associated outputs, produced by three well known handcrafted algorithms for active fire detection in the first part, and manually annotated masks in the second part. We also present a study on how different convolutional neural network architectures can be used to approximate these handcrafted algorithms, and how models trained on automatically segmented patches can be combined to achieve better performance than the original algorithms - with the best combination having 87.2% precision and 92.4% recall on our manually annotated dataset. The proposed dataset, source codes and trained models are available on Github (https://github.com/pereira-gha/activefire), creating opportunities for further advances in the field, 23 pages, 17 figures

With the rapid development of deep learning technology and improvement in computing capability, deep learning has been widely used in the field of hyperspectral image (HSI) classification. In general, deep learning models often contain many trainable parameters and require a massive number of labeled samples to achieve optimal performance. However, in regard to HSI classification, a large number of labeled samples is generally difficult to acquire due to the difficulty and time-consuming nature of manual labeling. Therefore, many research works focus on building a deep learning model for HSI classification with few labeled samples. In this article, we concentrate on this topic and provide a systematic review of the relevant literature. Specifically, the contributions of this paper are twofold. First, the research progress of related methods is categorized according to the learning paradigm, including transfer learning, active learning and few-shot learning. Second, a number of experiments with various state-of-the-art approaches has been carried out, and the results are summarized to reveal the potential research directions. More importantly, it is notable that although there is a vast gap between deep learning models (that usually need sufficient labeled samples) and the HSI scenario with few labeled samples, the issues of small-sample sets can be well characterized by fusion of deep learning methods and related techniques, such as transfer learning and a lightweight model. For reproducibility, the source codes of the methods assessed in the paper can be found at https://github.com/ShuGuoJ/HSI-Classification.git.

Surrogate-assisted evolutionary algorithms (SAEAs), which use efficient surrogate models or meta-models to approximate the fitness function in evolutionary algorithms (EAs), are effective and popular methods for solving computationally expensive optimization problems. During the past decades, a number of SAEAs have been proposed by combining different surrogate models and EAs. This paper dedicates to providing a more systematical review and comprehensive empirical study of surrogate models used in single-objective SAEAs. A new taxonomy of surrogate models in SAEAs for single-objective optimization is introduced in this paper. Surrogate models are classified into two major categories: absolute fitness models, which directly approximate the fitness function values of candidate solutions, and relative fitness models, which estimates the relative rank or preference of candidates rather than their fitness values. Then, the characteristics of different models are analyzed and compared by conducting a series of experiments in terms of time complexity (execution time), model accuracy, parameter influence, and the overall performance when used in EAs. The empirical results are helpful for researchers to select suitable surrogate models when designing SAEAs. Open research questions and future work are discussed at the end of the paper.

Knowledge graphs (KGs) are one of the most common frameworks for knowledge representation. However, they suffer from a severe scalability problem that hinders their usage. KG embedding aims to provide a solution to this issue. Nonetheless, general approaches are incapable of representing and reasoning about information not previously contained in the graph. This paper proposes to leverage semantic and ontological information for a significant benefit of knowledge graph completion, focusing on triple classification. The goal of this task is to determine whether a given fact holds. Furthermore, this paper also considers the classification of facts that include entities that have not been seen during training, denoted out-of-knowledge-base or OOKB entities. An incremental method is presented, composed of six stages. Although the proposal can be applied to any KG embedding model, this work focuses on its application for semantic matching models, such as ComplEx and DistMult. Compared to other approaches, our proposal is model-agnostic, computationally inexpensive, and does not require retraining. The results show that triple classification accuracy scales up to 15% with the proposed approach, as well as accelerating the convergence of the model to its optimal solution. Furthermore, facts containing OOKB entities can be classified with a reasonable accuracy.

Efficiency and robustness in remaining useful life (RUL) prediction are crucial in system health monitoring. Thus, the internal logic computation of a Deep LSTM model for RUL prediction is mainly shaped and evaluated over a training data-set and its performance examined on a testing data-set. This paper proposes a framework for testing robustness of deep Long Short Term Memory (LSTM) architecture for remaining useful life prediction that enables to gain confidence in the trained LSTM model for RUL prediction and ensures better quality. The resiliency of proposed Deep LSTM networks for RUL estimation using stress functions is first checked then the effect of the stress on model performance is analyzed. A comparison between the performance of the constructed mutant fuzzed Deep LSTM networks and the original Deep LSTM model for RUL prediction is provided to determine the quality of the RUL prediction model. Furthermore, the main purpose of this paper is to determine to what extent Deep LSTM models in the neighborhood of the trained LSTM model still have high test accuracy and quality scoring. Thus, the use of φ-stress operators shows that we could build stable and data-independent Deep LSTM models for RUL prediction. Finally, the proposed framework is validated using the Commercial Modular Aero Propulsion System Simulation (C-MAPSS) data-set.

Irregular time series data is becoming increasingly prevalent with the growth of multi-sensor systems as well as the continued use of unstructured manual data recording mechanisms. Irregular data and the resulting missing values severely limit the data's ability to be analysed and modelled for classification and forecasting tasks. Often, conventional methods used for handling time series data introduce bias and make strong assumptions on the underlying data generation process, which can lead to poor model predictions. Traditional machine learning and deep learning methods, although at the forefront of data modelling, are at best compromised by irregular time series data sets and fail to model the temporal irregularity of incomplete time series. Gated recurrent neural networks (RNN), such as LSTM and GRU, have had outstanding success in sequential modelling, and have been applied in many application fields, including natural language processing. These models have become an obvious choice for time series modelling and a promising tool for handling irregular time series data. RNNs have a unique ability to be adapted to make effective use of missing value patterns, time intervals and complex temporal dependencies in irregular univariate and multivariate time series data. In this paper, we provide a systematic review of recent studies in which gated recurrent neural networks have been successfully applied to irregular time series data for prediction tasks within several fields, including medical, human activity recognition, traffic monitoring and environmental monitoring. The review highlights the two common approaches for handling irregular time series data: missing value imputation at the data pre-processing stage and modification of algorithms to directly handle missing values in the learning process. Reviewed models are confined to those that can address issues with irregular time series data and does not cover the broader range of models that deal more generally with sequences and regular time series. This paper aims to present the most effective techniques emerging within this branch of research as well as to identify remaining challenges, so that researchers may build upon this platform of work towards further novel techniques for handling irregular time series data.

Experton theory, a generalization of probabilistic set theory, that is of great usefulness to group decision analysis, was first proposed as a means of improving the processing and analysis of opinions issued by experts. This theory produces an information-fusion mathematical object , the experton, which can be used in predictive problems to justify decisions based on well-constructed reasoning. The aim of this paper is to present an aggregative method of several expertons, with the idea that some of the groups of experts involved in producing these expertons may have more influence than others in the decision-making process. In this article, we carry out an aggregation analysis of expertons, not experts, which culminates in the creation of a new mathematical object. This object, which is called the weighted average multiexperton, is coherent with an experton-type object created from a weighting of the initial data provided by all experts. Since the aggregation method presented has been devised to represent the decision-maker’s attitude regarding the importance of different groups of experts, this approach represents a new tool for dealing with group decision-making problems. Additionally, the study presents some properties of the new object. Finally, the paper ends with an application for business decision-making.

Road landmark inventory is becoming an important data product for the maintenance of transport infrastructures. Several commercial sensors are available which include synchronized optical cameras that allowto build 360° panoramic images of the surroundings of the vehicle used for road inspection. This paper is devoted to the analysis of such panorama images,specifically the area that contains themost relevant information. Road lane landmark detection is posed as a two class classification problem that may be solved bymachine learningapproaches, such as Random Forest (RF) and ensembles of Extreme Learning Machines (V-ELM). Besides model parameter selection, a central problem is the construction of a labeled training and validation datasetto cope with the highly uncontrolled conditions of image capture. Besides, human labor cost makes image data labeling a very expensive process. This paper proposes an open ended Active Learning (AL) approach involving a human oraclein the loop who provides the data labeling and can trigger the AL process when detection quality is degraded by the change in imaging conditions. The paper reports encouraging results over a collection of sample images selected from an industrial road landmark inventory operation. As an additional contribution, this paper assesses the ability of AL to overcomesome of the issues raised by highly class imbalanced datasets.

Control performance assessment (CPA) is vital to ensure the safety of control systems. However, most multivariate CPA methods are limited to the system with explicit knowledge. Recently, it has been recognized that high predictability of closed-loop outputs implies poor control performance. This paper proposes a probabilistic CPA (PCPA), which is compatible with the above promising idea. This paper constructs a modified Gaussian process regression (GPR) model to quantitatively​ estimate the prediction uncertainty of outputs with only routine closed-loop data as input and focuses on the prediction variance of interest. As a non-parametric and probabilistic method, the proposed framework can handle the nonlinearity and random uncertainties inherent in complex control systems. Combined with the varying window size strategy, a novel performance metric, called disruption resistance (DR) here, is designed to characterize different control performance. The evaluation confidence and uncertainty can be revealed with concurrent analytics of GPR and Gaussian process classification when performing the performance assessment. This gives rise to a reliable and pragmatic PCPA framework, which shows more accurate and comprehensive results in the application to both simulated and real industrial processes.

Depth estimation is a classic task in computer vision, which is of great significance for many applications such as augmented reality, target tracking and autonomous driving. Traditional monocular depth estimation methods are based on depth cues for depth prediction with strict requirements, e.g. shape-from-focus/ defocus methods require low depth of field on the scenes and images. Recently, a large body of deep learning methods have been proposed and has shown great promise in handling the traditional ill-posed problem. This paper aims to review the state-of-the-art development in deep learning-based monocular depth estimation. We give an overview of published papers between 2014 and 2020 in terms of training manners and task types. We firstly summarize the deep learning models for monocular depth estimation. Secondly, we categorize various deep learning-based methods in monocular depth estimation. Thirdly, we introduce the publicly available dataset and the evaluation metrics. And we also analysis the properties of these methods and compare their performance. Finally, we highlight the challenges in order to inform the future research directions.

With the advent of smart phones, the popularity of free Android applications has risen rapidly. This has led to malicious Android apps being involuntarily installed, which violate the user privacy or conduct attack. Malware detection on Android platforms therefore is a growing concern because of the undesirable similarity between malicious behavior and benign behavior, which can lead to slow detection, and allow compromises to persist for comparatively long periods of time in infected phones. The contributions of this paper are first a multiple dimensional, kernel feature-based framework and feature weight-based detection (WBD) designed to categorize and comprehend the characteristics of Android malware and benign apps. Furthermore, our software agent is orchestrated and implemented for the data collection and storage to scan thousands of benign and malicious apps automatically. We examine 112 kernel attributes of executing the task data structure in the Android system and evaluate the detection accuracy with a number of datasets of various dimensions. We find that memory- and signal-related features contribute to more precise classification than schedule-related and other descriptors of task states listed in our paper. Particularly, memory-related features provide fine-grain classification policies for preserving higher classification precision than the signal-related and others. Furthermore, we study and evaluate 80 newly infected attributes of the Android kernel task structure, prioritizing the 70 features of most significance based on dimensional reduction to optimize the efficiency of high-dimensional classification. Our second contribution is that our experiments demonstrate that, as compared to existing techniques with a short list of task structure features (16 or 32 features), our method can achieve 94%-98% accuracy and 2%–7% false positive rate, while detecting malware apps with reduced-dimensional features that adequately abbreviate online malware detections and advance offline malware inspections.

While meta learning approaches have achieved remarkable success, obtaining a stable and unbiased meta-learner remains a significant challenge, since the initial model of a meta-learner could be too biased towards existing tasks to adapt to new tasks. In order to avoid a biased meta-learner and improve its generalizability, this paper proposes a generic meta learning method that aims to learn an unbiased meta-learner towards a variety of tasks before its initial model is adapted to unseen tasks. Specifically, this paper presents a meta weight learning method for minimizing the inequality of performance across different training tasks. An end-to-end training approach is introduced for the proposed algorithm that allows for effectively learning weight and initializing the network model. Alternatively, a variety of measurement methods of weight is also designed to test the effectiveness of different weight learning methods on the improvement of model-agnostic meta-learning algorithm. The simulation results show that the proposed meta weight learning method not only outperforms state-of-the-art meta learning algorithms, but also is superior to other manually designed measurement methods of weight on discrete and continuous control problems.

The structure and functionality of the POLEMAD database constructed on the basis of a study using Electromagnetic Articulograph AG 500, an acoustic camera, and 3 video cameras are described in the paper. The article describes also data types stored in the database including speaker data, EMA data, video and sound recordings, phonetic information, and dynamic Bayesian network (DBN) models. The database allows for selective extraction of various types of samples for further analysis, which is performed by SQL queries generated in MATLAB® using Database Toolbox™. The possibilities of potential future application of the database in statistical analysis and automation of experiments on speech inversion using DBN are described in the paper as well.

Internet of things architecture is the integration of real-world objects and places with the internet. This booming in technology is bringing ease in our lifestyle and making formerly impossible things possible. Internet of things playing a vital role in bridging this gap easily and rapidly. IoT is changing our lifestyle and the way of working the technologies, by bringing them together at the one page in several application areas of daily life. However, IoT has to face several challenges in the form of cyber scams, one of the major challenges IoT has to face is the likelihood of Ransomware attack. Ransomware is a malicious kind of software that restricts access to vital information in some way and demand payment for getting access to this information. The ransomware attack is becoming widespread daily, and it is bringing disastrous consequences, including loss of sensitive data, loss of productivity, data destruction, and loss of reputation and business downtime. Which further leads to millions of dollar daily losses due to the downtime. This is inevitable for organizations to revise their annual cybersecurity goals and need to implement proper resilience and recovery plan to keep business running. However, before proceeding towards providing a practical solution, there is a need to synthesize the existing data and statistics about this crucial attack to make aware to the researchers and practitioners. To fill this gap, this paper provides a comprehensive survey on evolution, prevention and mitigation of Ransomware in IoT context. This paper differs from existing in various dimensions: firstly, it provides deeper insights about Ransomware evolution in IoT. Secondly; it discusses diverse aspects of Ransomware attacks on IoT which include, various types of Ransomware, Current research in Ransomware, Existing techniques to prevent and mitigate Ransomware attacks in IoT along with the ways to deal with an affected machine, the decision about paying the ransom or not, and future emerging trends of Ransomware propagation in IoT. Thirdly, a summary of current research is also provided to show various directions of research. In sum, this detailed survey is expected to be useful for researchers and practitioners who are involved in developing solutions for IoT security.

As a new type of network, the types of attack in the Internet of Vehicles (IoV) are constantly emerging and changing. Consequently, the machine learning based intrusion detection model has to update to cope with new attacks. However, existing machine learning based IoV intrusion detection schemes require large amounts of labeled data to complete model updates. For new attacks, the IoV cloud is also difficult to identify in time, which requires a lot of labor and time cost in IoV. To solve above issue, this paper employs transfer learning and proposes two model update schemes based on whether the IoV cloud can timely provide a small amount of labeled data for a new attack. The first one is the cloud-assisted update scheme where the IoV cloud can provide a small amount of data. And the second one is the local update scheme where the IoV cloud cannot provide any labeled data timely. In this paper, the local update scheme obtains pseudo label of the unlabeled data in new attacks via pre-classifies and uses the pseudo-labeled data for multiple rounds of transfer learning. Then the vehicle can complete the update without obtaining any labeled data through the IoV cloud. The experimental results show that compared with the existing method, our two schemes have improved the detection accuracy by at least 23%.

Acute hypotensive episode (AHE) is a common serious postoperative complication in ICU, which may raise multiple system failure (especially of cardiac and respiratory kinds), and even cause death. Timely and effective clinical intervention is obviously vital to the saving of patients. AHE detection involves physiological time-series monitoring, processing and prediction technologies, which can offer insights to neuroscientists, biologists, and even provide support for clinicians. This paper presents a hybrid artificial intelligence model combined with CEEMDAN (complete ensemble empirical mode decomposition with adaptive noise, a typical method for physiological signal decomposition), deep learning, multiple gene expression programming and fuzzy expert system for AHE detection. In this paper, the physiological signal is selected from a benchmark dataset, for example MIMIC-II (Multiparameter Intelligent Monitoring in Intensive Care II), which collects large scale real patients’ data for clinical research. In the hybrid model, a typical signal decomposition method is employed for AHE signal processing, and an autoencoder based deep neural network is established for feature extraction. Finally, a reliable and explainable classifier is presented by fusing gene expression programming and the fuzzy method. Experimental results based on real data set demonstrate that the proposed method outperforms state-of-the-art AHE detection methods by achieving the prediction accuracy of 88.14% in 2866 records.

The natural concept ontology structure of clothes has enabled easy management of large quantities of fashion images for online retailers and it is meaningful to study how to automatically recognize fashion images for both commercial promotion and academic research. In this paper, a new hierarchical approach is developed for large-scale fashion recognition. We first embed concept ontology into deep convolutional neural network (CNN) by adopting multiple deep CNN branches to learn node-specific features and classifiers explicitly. Then, we introduce a hierarchical knowledge distillation method to further improve the performance of fashion recognition. Finally, we employ the proposed approach for fashion recommendation. To deal with hierarchical deep learning constrains, we leverage back propagation to simultaneously refine the shared deep CNNs and the diverse CNN branches for relevant node features and classifiers by using our joint objective function. The main advantages of this paper lie in (1) providing an effective way for recognizing rich semantic explanations of fashion images without training large or multiple networks, and (2) saving the storage&time costs by learning personalized features and classifiers for each tree node. The experimental results on both our organized fashion dataset and the public DeepFashion dataset have verified the effectiveness and efficiency of the proposed approach on both hierarchical fashion recognition and within category fine-grained fashion recommendation.

Highlights • The minimum cost consensus model is proposed to coordinate the opinions of experts. • A new method of LSGDM clustering is proposed to improve CRP efficiency. • A consensus index CI reflecting the harmony degree between experts is devised. • A robust counterpart of robust optimization consensus is implemented., Recently, large-scale group decision making (LSGDM) in social network comes into being. In the practical consensus of LSGDM, the unit adjustment cost of experts is difficult to obtain and may be uncertain. Therefore, the purpose of this paper is to propose a consensus model based on robust optimization. This paper focuses on LSGDM, considering the social relationship between experts. In the presented model, an expert clustering method, combining trust degree and relationship strength, is used to classify experts with similar opinions into subgroups. A consensus index, reflecting the harmony degree between experts, is devised to measure the consensus level among experts. Then, a minimum cost model based on robust optimization is proposed to solve the robust optimization consensus problem. Subsequently, a detailed consensus feedback adjustment is presented. Finally, a case study and comparative analysis are provided to verify the validity and advantage of the proposed method.

This paper addresses the access control issue for the comprehensive and distributed Internet of Things (IoT) environments. The typical eXtensible Access Control Markup Language (XACML) which implements the sophisticated access conditions in XML files, is widely used to guarantee access control decisions for the distributed IoT environments. To the best of our knowledge, the typical XACML-based access control schemes never consider the authentication run-time parameters. Moreover, the access control schemes that are mainly based on the typical XACML cannot secure themselves against some kinds of attacks, such as the Masquerade attack. Also, those schemes are not secure in opposition to Man-in-the-Middle (MITM) attack. Therefore, this paper proposes an adaptive XACML scheme that extends the typical XACML by integrating an access code generation and verification schemes for heterogeneous distributed IoT environments. Our adaptive XACML scheme considers some sensitive authentication run-time parameters before authorizing the user. Moreover, our scheme is proven secure against the Masquerade attack and MITM attack through hashing the generated access code using Message Digest Algorithm-5 (MD5) and Secure Hash Algorithm-1 (SHA-1) Checksum Utility. The experimental analysis of many different configurations supports the efficacy and efficiency of our adaptive XACML. It also shows exceptional compatibility and performance with different implementations. The processing time comparison between our adaptive XACML and Typical XACML, has shown that there is a low time overhead when using our adaptive XACML. This processing time overhead is nothing compared to the extra features that have been achieved in excess of the typical XACML and the security against Masquerade and MITM attacks. Therefore, our adaptive XACML scheme has the capability to be applied in various distinct distributed environments not only IoT.

Considering that most popular models solving aspect-level sentiment classification problems mainly focus on designing complicated neural networks to scale the importance of each word in the sentence, this paper addresses this problem from the view of semantic space. Motivated by the fact that the senses of a word can be sophisticatedly embedded into the semantic space using a distributed representation, this paper hypothesizes that each sense of a word can be represented by one or more specific dimensions, and thus the target of aspect-level sentiment classification can be simplified into searching the related dimensions for the aspects and sentiments concerned. Particularly, an Attention Vector (ATV) based on attention mechanism is designed for each aspect in terms of a specific task, which involves two sub-vectors, i.e., a Dimension Attention Vector (DATV) and a Sentiment Attention Vector (SATV). The DATV determines the significances of different dimensions based on their correlations with an aspect; and the SATV allocates weights for the attributes of words, which are decided by sentiment polarities and part-of-speech (PoS) tagging. Given a sub-dataset related to an aspect, the ATV will be optimized by an Artificial Bee Colony (ABC) algorithm with a Support Vector Machine (SVM) classifier, the objective of which is to maximize classification accuracy. Intrinsically, the DATV can reduce the ambiguity existing in polysemy, meanwhile, the SATV is an auxiliary means for the optimization of the DATV, which can help eliminate the misunderstandings caused by antonyms. Then, the optimized DATV will be applied on a Convolutional Neural Network (CNN) model via simply scaling the pretrained word embeddings as inputs (named as ATV-CNN model). Experimental results show that the ATV-CNN model can have substantial advantages when compared with state-of-the-art models.

Analyzing the runtime behaviors of Android apps is crucial for malware detection. In this paper, we attempt to learn the behavior level features of an app from function calls. The challenges of this task are twofold. First, the absence of function attributes hinders the understanding of app behaviors. Second, the graphical representation of function calls cannot be directly processed by classical machine learning algorithms. In this paper, we develop two methods to overcome these challenges. Based on function embedding, we first propose the concept of enhanced function call graphs (E-FCGs) to characterize app runtime behaviors. We then develop a Graph Convolutional Network (GCN) based algorithm to obtain vector representations of E-FCGs. Extensive experiments show that the features learned by our method can achieve surprisingly high detection performance on a variety of classifiers (e.g., LR, DT, SVM, KNN, RF, MLP and CNN), significantly outperforming the traditional static features.

Plants are susceptive to various diseases in their growing phases. Early detection of diseases in plants is one of the most challenging problems in agriculture. If the diseases are not identified in the early stages, then they may adversely affect the total yield, resulting in a decrease in the farmers' profits. To overcome this problem, many researchers have presented different state-of-the-art systems based on Deep Learning and Machine Learning approaches. However, most of these systems either use millions of training parameters or have low classification accuracies. This paper proposes a novel hybrid model based on Convolutional Autoencoder (CAE) network and Convolutional Neural Network (CNN) for automatic plant disease detection. To the best of our knowledge, a hybrid system based on CAE and CNN to detect plant diseases automatically has not been proposed in any state-of-the-art systems present in the literature. In this work, the proposed hybrid model is applied to detect Bacterial Spot disease present in peach plants using their leaf images, however, it can be used for any plant disease detection. The experiments performed in this paper use a publicly available dataset named PlantVillage to get the leaf images of peach plants. The proposed system achieves 99.35% training accuracy and 98.38% testing accuracy using only 9,914 training parameters. The proposed hybrid model requires lesser number of training parameters as compared to other approaches existing in the literature. This, in turn, significantly decreases the time required to train the model for automatic plant disease detection and the time required to identify the disease in plants using the trained model.

In this paper, we propose a method for obtaining sentence-level embeddings. While the problem of securing word-level embeddings is very well studied, we propose a novel method for obtaining sentence-level embeddings. This is obtained by a simple method in the context of solving the paraphrase generation task. If we use a sequential encoder-decoder model for generating paraphrase, we would like the generated paraphrase to be semantically close to the original sentence. One way to ensure this is by adding constraints for true paraphrase embeddings to be close and unrelated paraphrase candidate sentence embeddings to be far. This is ensured by using a sequential pair-wise discriminator that shares weights with the encoder that is trained with a suitable loss function. Our loss function penalizes paraphrase sentence embedding distances from being too large. This loss is used in combination with a sequential encoder-decoder network. We also validated our method by evaluating the obtained embeddings for a sentiment analysis task. The proposed method results in semantic embeddings and outperforms the state-of-the-art on the paraphrase generation and sentiment analysis task on standard datasets. These results are also shown to be statistically significant., Comment: This work is an extension of our COLING-2018 paper arXiv:1806.00807