Search

The growing number of scientific papers and document sources underscores the need for methods capable of evaluating the quality of publications. Researchers who are looking for relevant papers for their studies need ways to assess the scientific value of these documents. One approach involves using semantic search engines that can automatically extract important knowledge from the growing body of text. In this study, we introduce a new metric called "MAATrica," which serves as the foundation for an innovative method designed to evaluate research papers. MAATrica offers a new way to analyze and categorize text, focusing on the consistency of research documents in the life sciences, particularly in the fields of medicinal and nutraceutical chemistry. This method utilizes semantic descriptions to cover in silico experiments, as well as in vitro and in vivo essays. Created to aid in evaluation processes like peer review, MAATrica uses toolkits and semantic applications to build the proposed measure, identify scientific entities, and gather information. We have applied MAATrica to roughly 90,000 papers and present our findings here. [Display omitted] • MAATrica is a novel metric for assessing the coherence of methodologies in research papers within the fields of medicinal and nutraceutical chemistry. • MAATrica utilizes SciWalker (SW) as semantic search engine and ontologies to automate knowledge extraction and evaluate research papers. • MAATrica metric has been tested and validated using a dataset comprising approximately 90,000 papers within the SW platform. • MAATrica's reliability has undergone testing through comparisons with manual evaluations, revealing strong agreement and potential support for peer review. • MAATrica employs a user-controlled and customizable ontology, enabling personalized analysis of research papers in the fields of medicinal and nutraceutical chemistry. [ABSTRACT FROM AUTHOR]

The fact-condition statement plays a significant role in the geological papers, where geological theory or phenomenon as a fact is constrained by its precondition. This constraint relation conveys a lot of key information in geological papers, but it has been ignored in most of studies on the information extraction. Although the rule-based algorithms have been proposed in the previous studies to extract the relationship between facts and conditions, but their low accuracy limit usability. Therefore, this paper introduces a series of mechanisms and methods to improve the extraction accuracy. We firstly utilize three kinds of pooling for extracting the semantic features of fact-condition tuples to find that balanced semantic representations can more accurately reflect their relations. In addition, semantic correlation between facts and conditions is also designed to mine their relations via the measurement of semantic similarity or semantic distance. Based on the above, the Deep Structured Semantic Model with Multi-Grained Representation (MGR-DSSM) is proposed. In this model, the Deep Structured Semantic Model (DSSM) is utilized two individual projections for denoising semantic correlation between facts and conditions, while the Multi-Grained Representation (MGR) is designed for encoding their semantic correlation at multiple granularities. Furthermore, the linear MGR-DSSM with SciBERT achieves the best result with 70.41% of accuracy and 74.55% of F1 score, obviously outperforming other baseline models. Finally, the linear MGR-DSSM with SciBERT is adopted to extract super relation from 722 geological paper abstracts, and the geological knowledge graph with fact-condition statement is built. With more accurate relation between fact and condition, the information validity of geological knowledge graph is enhanced. • The relation between facts and conditions play an essential role in the geological paper, the extraction of which is studied for the first time. • A series of mechanisms are proposed to improve the accuracy of fact-condition relation extraction. • The MGR-DSSM is proposed, which obviously outperforms other baselines in super relation extraction. [ABSTRACT FROM AUTHOR]

This paper explores the utilization of randomized SVD (rSVD) in the context of kernel matrices arising from radial basis functions (RBFs) for the purpose of solving interpolation and Poisson problems. We propose a truncated version of rSVD, called trSVD, which yields a stable solution with a reduced condition number in comparison to the non-truncated variant, particularly when manipulating the scale or shape parameter of RBFs. Notably, trSVD exhibits exceptional proficiency in capturing the most significant singular values, enabling the extraction of critical information from the data. When compared to the conventional truncated SVD (tSVD), trSVD achieves comparable accuracy while demonstrating improved efficiency. Furthermore, we explore the potential of trSVD by employing scale parameter strategies, such as leave-one-out cross-validation and effective condition number. Then, we apply trSVD to solve a 2D Poisson equation, thereby showcasing its efficacy in handling partial differential equations. In summary, this study offers an efficient and accurate solver for RBF problems, demonstrating its practical applicability. The code implementation is provided to the scientific community for their access and reference. [ABSTRACT FROM AUTHOR]

Access points to Heliophysics information are often poorly inter-linked to one another, making it challenging for researchers to integrate information, mature their understanding, and incorporate more complex ideas and relationships into their analyses. The authors reason this behavior to be a direct conseaquence of the lack of infrastructure for knowledge and data discovery in Heliophysics. Building an online library resource for Heliophysics addresses this gap in resources, and can also address other troublesome gaps related to community-building, such as access to help from other scientists, example analysis tutorials, and collaboration opportunities. This paper envisions a resource that better connects existing Heliophysics information and knowledge sources, the LIbrary for KnowledgE and Discovery (LIKED), which will be an online resource that interlinks information across databases (e.g., observational and publication), researcher interactions (e.g., conferences and discussion boards), and educational materials. The result is a rich and searchable environment to provide users a more productive experience for research and collaboration. In this paper, we provide the conceptual model for LIKED. We expand to detail how LIKED will serve to connect the various knowledge and research components available internationally together in an uniformly searchable manner. Finally, this work outlines various contributions the community can make towards this goal. [ABSTRACT FROM AUTHOR]

Geographic spaces is a vast and complex system involving multiple elements and nonlinear interactions of these elements, and rich in geographical phenomena, processes and patterns. Artificial intelligence methods (AI) are increasingly utilized to extract information of interest, patterns and insights from massive remote sensing (RS) images. Among them, two representative paradigms for RS interpretation are knowledge-driven symbolism and data-driven connectionism. Knowledge-driven approaches are certain and theoretically sound, which consider the regularities, formulae with accompanying theorems, and expert knowledge. However, it is difficult to address huge-scale RS data and complex nonlinear problems. The data-driven paradigm has stronger data learning and representation ability, yet poor interpretability, low trustworthiness of interpretation results, and reliance on massive labeled data. To address these limitations, we argue that current RS intelligent interpretation requires the guidance of geographic ideas and a unified theoretical framework rather than the independent use of remotely sensed data, duplicative development of interpretation models, or isolated applications of geographic information extraction. Thus, we introduce the concepts of trustworthy RS interpretation (TRSI) with multi-granularity of space, time, and attribute to facilitate the understanding of the morphology, types, and indicators (visual perception) and the perspective of the structure, evolution, and trends (geographical cognition) of complex geographical space. Second, integrated technologies of the TRSI with multiple scales and stages are developed, including a pixel-level visual perception, a geo-parcel-level quantitative inversion, and a scene-level geographical cognition. Finally, a RS big data interpretation system with quantitative semantic parsing is designed to support precise applications with cloud–edge-end collaboration. Rather than a simple overview and accumulation of traditional RS intelligent interpretation methods, this paper analyzes current challenges in RS interpretation and proposes new concepts, technologies, and applications of TRSI. It will empower us to perceive the complex geographical space comprehensibly and reveal its structural composition and evolutionary mechanism, serving geospatial decision applications such as evaluation, planning, and forecasting. [ABSTRACT FROM AUTHOR]

Real-time semantic segmentation, as a key technique for scene understanding, has been an important research topic in the field of computer vision in recent years. However, existing models are unable to achieve good segmentation accuracy on mobile devices due to their huge computational overhead, which makes it difficult to meet actual industrial requirements. To address the problems faced by current semantic segmentation tasks, this paper proposes a lightweight bilateral asymmetric residual network (LBARNet) for real-time semantic segmentation. First, we propose the bilateral asymmetric residual (BAR) module. This module learns multi-scale feature representations with strong semantic information at different stages of the semantic information extraction branch, thus improving pixel classification performance. Secondly, the spatial information extraction (SIE) module is constructed in the spatial detail extraction branch to capture multi-level local features of the shallow network to compensate for the lost geometric information in the downsampling stage. At the same time, we design the attention mechanism perception (AMP) module in the jump connection part to enhance the contextual representation. Finally, we design the dual branch feature fusion (DBF) module to exploit the correspondence between higher-order features and lower-order features to fuse spatial and semantic information appropriately. The experimental results show that LBARNet, without any pre-training and pre-processing and using only 0.6M parameters, achieves 70.8% mloU and 67.2% mloU on the Cityscapes dataset and Camvid dataset, respectively. LBARNet maintain a high segmentation accuracy while using a smaller number of parameters compared to most existing state-of-the-art models. [Display omitted] • This paper proposes a Bilateral Asymmetric Residual (BAR) module, a Spatial Information Extraction (SIE) module, an Attention Mechanism Perception (AMP) module and a Dual Branch Feature Fusion (DBF) module. • A Lightweight Bilateral Asymmetric Residual Network (LBARNet) for real-time image semantic segmentation is proposed in this article. • The experimental results show that LBARNet achieves 70.8% and 67.2% segmentation accuracy on two challenging datasets (Cityscapes and CamVid) using only 0.6M parameters. [ABSTRACT FROM AUTHOR]

Multi-label classification has recently attracted research interest as a data mining task. Many current applications in data mining address problems that have instances belonging to more than one class. This requires the development of new efficient methods. Instance selection has been used in multi-label learning to improve the execution time and classification performance of many learning methods. Following the single-label approach, instance selection has been applied by selecting or unselecting the same instances for all labels. In this paper, we present a different and novel approach. An instance might be useful for some labels and harmful for others; therefore, our algorithm allows each instance to be discarded, selected, or only partially selected for use in the classification of certain labels. An extensive comparison using 45 datasets shows the usefulness of our approach in improving the current instance selection methods for multi-label problems, as well as the ability of our algorithm to compete with other more complex multi-label classification methods. • This paper presents a new framework for instance selection for multi-label datasets. • The method improves several state-of-the-art methods. • The method is scalable to large multi-label datasets. • Interesting new research lines are opened by the framework. [ABSTRACT FROM AUTHOR]

The Black Hole Clustering (BHC) algorithm is a density-based partitional clustering method inspired by the Density-based Spatial Clustering of Applications with Noise (DBSCAN). It does not require the number of clusters nor the computation of the pair-wise distance matrix between the data points, making it faster than DBSCAN. Also, it only needs one parameter that is intuitively easier to set than the epsilon parameter of DBSCAN. However, BHC needs the allocation of the so-called black holes that have to be linearly independent, making the algorithm in its current version suitable only for two or three-dimensional data sets. In this paper, we propose a generalized version of the black hole clustering algorithm (GBHC) by introducing a novel black hole allocation procedure for higher-dimensional data spaces. Furthermore, the proposed method is data-independent, so we have to run it once to obtain the black hole positions for all finite-dimensional metric spaces. We performed extensive computational experiments to compare GBHC with DBSCAN. The results show that both algorithms obtain comparable clustering solutions. GBHC, however, outperforms DBSCAN in computational complexity and explainability. • We propose a novel method to place the black holes for high-dimensional data sets. • This method is called Generalized Black Hole Clustering and it is data-independent. • We run the new method once to get black hole positions in all finite metric spaces. • We compare the GBHC and DBSCAN algorithms using several validation measures. [ABSTRACT FROM AUTHOR]

Due to popular successes (e.g., ChatGPT) Artificial Intelligence (AI) is on everyone's lips today. When advances in biotechnology are combined with advances in AI unprecedented new potential solutions become available. This can help with many global problems and contribute to important Sustainability Development Goals. Current examples include Food Security, Health and Well-being, Clean Water, Clean Energy, Responsible Consumption and Production, Climate Action, Life below Water, or protect, restore and promote sustainable use of terrestrial ecosystems, sustainably manage forests, combat desertification, and halt and reverse land degradation and halt biodiversity loss. AI is ubiquitous in the life sciences today. Topics include a wide range from machine learning and Big Data analytics, knowledge discovery and data mining, biomedical ontologies, knowledge-based reasoning, natural language processing, decision support and reasoning under uncertainty, temporal and spatial representation and inference, and methodological aspects of explainable AI (XAI) with applications of biotechnology. In this pre-Editorial paper, we provide an overview of open research issues and challenges for each of the topics addressed in this special issue. Potential authors can directly use this as a guideline for developing their paper. • This pre-Editorial provides an overview of some hot topics in AI for life. • The benefit for the reader is to get a short, crisp, concise intro into AI. • Authors interested in this Special Issue will get a scope of topics. • The take away message is that in the future no biotechnology can do without AI. [ABSTRACT FROM AUTHOR]

Lower-dimensional representation of scientific text has attracted much attention among researchers due to its impact on many data mining and recommendation tasks. This paper studies two main research streams in scientific literature representation. First, both local and distributed representation viewpoints are reviewed and their advantages and disadvantages in lower dimensional representation are discussed. The paper then proposes a novel hybrid distributed technique for text representation. Using scientific articles as the major source of textual information, both the article's content and citation network are used to build a distributed and universal lower dimensional representation. The superiority of the new technique to the traditional methods is then justified in predicting the existence of links in large citation graphs. [ABSTRACT FROM AUTHOR]

Accurate and effective access to agricultural and pastoral information is of great practical importance for scientific planning in agriculture and pastoral husbandry and food security reserves. This paper utilised intra-annual monthly synthetic data from Sentinel-1 and Sentinel-2, combined with band information and index information to simultaneously obtain intra-annual phenological change information and extracted different types of information for random forest classification, thus obtaining the fine-grained agricultural and pastoral map of Jingyang District. On this basis, the effects of different data sources and the importance of different features on the classification results were further analysed. The results show that in the classification results of the Sentinel-1 data method, March feature is the most important and May feature has the least influence on the classification results. In the classification results of the Sentinel-2 data method and the combined Sentinel-1 data and Sentinel-2 data, the December feature has the greatest importance and the February feature has the least impact on the classification results. The NDVI index has the highest importance and the EVI index has the lowest importance score. The importance of band features varies in the classification results of different data sources. The accuracy of the agricultural and pastoral information in Jingyang District is 83.31% and the kappa is 0.76. Rice is the largest crop with an area of 239,830,800 m2, mainly located in the northwest and mountainous areas of Jingyang District. It is followed by green manure with an area of 60,274,000 m2, mainly in the mountainous part. This study demonstrates the feasibility of extracting agricultural and pastoral information from remote sensing images, and also provides basic data for future agricultural and pastoral planning in Jingyang District. [ABSTRACT FROM AUTHOR]

This paper reviews the recent progress of small angle scattering (SAS) techniques, mainly including X-ray small angle scattering technique (SAXS) and neutron small angle scattering (SANS) technique, in the study of metal-organic framework (MOF) colloidal materials (CMOFs). First, we introduce the application research of SAXS technique in pristine MOFs materials, and review the studies on synthesis mechanism of MOF materials, the pore structures and fractal characteristics, as well as the spatial distribution and morphological evolution of foreign molecules in MOF composites and MOF-derived materials. Then, the applications of SANS technique in MOFs are summarized, with emphasis on SANS data processing method, structure modeling and quantitative structural information extraction. Finally, the characteristics and developments of SAS techniques are commented and prospected. It can be found that most studies on MOF materials with SAS techniques focus mainly on nanoporous structure characterization and the evolution of pore structures, or the spatial distribution of other foreign molecules loaded in MOFs. Indeed, SAS techniques take an irreplaceable role in revealing the structure and evolution of nanopores in CMOFs. We expect that this paper will help to understand the research status of SAS techniques on MOF materials and better to apply SAS techniques to conduct further research on MOF and related materials. [Display omitted] [ABSTRACT FROM AUTHOR]

Fusion of information from infrared and visible images from different modalities has always been a hot research topic in the field of image processing. In order to solve the problem that existing fusion methods cannot fully utilize the deep information of images, this paper proposes a fusion algorithm DAE-Nest based on depth information extraction and enhancement. The algorithm consists of encoder-feature enhancement-fusion layer-decoder modules for unsupervised learning. At first, the improved coding structure is used to extract multi-scale features of the source image; secondly, the extracted features are data enhanced to fully utilize the deep image information; then, the enhanced features are fused according to the corresponding fusion strategy; finally, the decoder takes the input of fused features and reconstructs them into an information-rich fused image. DAE-Nest was subjected to corresponding fusion test experiments on the TNO and RoadScene data sets respectively and compared with 9 excellent fusion algorithms. In terms of subjective evaluation, DAE-Nest can make full use of the deep information in the source image, and the generated fusion image also conforms to the visual perception of the human eyes; in terms of objective evaluation indicators, DAE-Nest's average gradient and spatial frequency on the TNO data set, it achieved the best performance in the three evaluation indicators of structural similarity and structural similarity, and also achieved the best performance in the fusion quality evaluation index on the RoadScene data set. Therefore, the fusion method proposed in this paper has achieved excellent results both subjectively and objectively, and can effectively fuse infrared and visible images, laying a good foundation for subsequent visual tasks. • Design a deep multi-scale feature extraction structure to fully extract the deep scale features of the image. • Design feature enhancement structures to make full use of the deep scale features of the image. • Design end-to-end integration modules IRS (Inception ResNet Shallow) and IRD (Inception ResNet Deep). • The proposed method was evaluated on two test sets and good fusion results were achieved. [ABSTRACT FROM AUTHOR]

Existing deep image classification techniques strive to suppress data uncertainty for various reasons such as blur, occlusion, noise, and label error and advance to higher accuracy. However, they ignore data uncertainty-induced imprecision and thus do not work as intended. In this paper, we propose a deep open-source framework to mine and reason the imprecision of such training and test sets, which can present better performance and reduce the risk of misclassification. First, we design a label reassignment mechanism. It allows the network to reassign training labels and allow imperfect training samples with multiple labels. As a result, they are removed from the original classes and considered new imprecise samples to represent partial ignorance. Second, we propose a new imbalanced data enhancement architecture to learn a generalized representation of each (precise and imprecise) class. It helps the network fuse the auxiliary information from both precise and imprecise classes, which is beneficial to extract more distinctive class features from single-labeled samples and characterize uncertainty-induced imprecision in the test set by imprecise test samples. Afterward, methodological analyses and empirical evaluations are conducted. The proposed framework is demonstrated to present better performance on different typical networks (Resnet50, MobileNetV2, DenseNet121, EfficientNetB0, ShuffleNetV2, SENet, SqueezeNet, Xception) based on five publicly available datasets (Imagewoof-5, Flowers, Monkey, Butterfly, and Cifar-10). In addition, several targeted deep techniques for uncertain images or imprecise results are also employed as comparisons to prove the superiority of the proposed framework. • We develop a deep open-source framework to mine and reason imprecise information. • We provide a multi-label reassignment mechanism to select imprecise samples. • We design an imbalanced data enhancement architecture to fuse different information. • The method can expand training samples and reduce the gap of different classes. [ABSTRACT FROM AUTHOR]

• Pipeline to extract information from historical tables with different layouts. • Two information extraction systems are compared at each phase of the pipeline. • Deal with multispan and classify headers by content. • Improvements in both systems achieve new state-of-the-art results in HisClima tables. Document Image Understanding is a demanding Pattern Recognition problem that requires complex recognition models. This problem is even more difficult for document images with complicated layouts like tables, where the reading order is often intrinsically ambiguous, and consequently, the context is generally ambiguous as well. In this paper, we compare two machine learning approaches for extracting information in pre-printed historical tables with handwritten information. We analyze the performance of each approach at each step of the extraction process over different corpora, up to a realistic scenario where documents with different table layouts written by different hands are used. The results are good in general and show that a model based on Multilayer Perceptrons yields better results on more homogeneous documents, while another model based on Graph Neural Networks generalizes better on heterogeneous corpora. [ABSTRACT FROM AUTHOR]

• A word embedding-based topic model (WETM) for short text documents. • Sparsity problem removed in short text and discovered structural information for topics and words. • A modified collapsed Gibbs sampling algorithm to find the parameters for WETM. • WETM achieved better classification, topic coherence, topic quality, and clustering results. • The execution time is lower for WETM as compared to baseline topic models. Short texts are a common source of knowledge, and the extraction of such valuable information is beneficial for several purposes. Traditional topic models are incapable of analyzing the internal structural information of topics. They are mostly based on the co-occurrence of words at the document level and are often unable to extract semantically relevant topics from short text datasets due to their limited length. Although some traditional topic models are sensitive to word order due to the strong sparsity of data, they do not perform well on short texts. In this paper, we propose a novel word embedding-based topic model (WETM) for short text documents to discover the structural information of topics and words and eliminate the sparsity problem. Moreover, a modified collapsed Gibbs sampling algorithm is proposed to strengthen the semantic coherence of topics in short texts. WETM extracts semantically coherent topics from short texts and finds relationships between words. Extensive experimental results on two real-world datasets show that WETM achieves better topic quality, topic coherence, classification, and clustering results. WETM also requires less execution time compared to traditional topic models. [ABSTRACT FROM AUTHOR]

This paper describes the performances of a marine SO 2 absorption scrubber installed onboard a large Ro-Ro cargo ship. The study is based on the reconstruction of an extensive dataset from one-year continuous monitoring of the scrubber's performances and operating conditions. The dataset has been interpreted with a conventional analytical, physical-mathematical, model for absorbers' rating and its combination with an Artificial Intelligence (AI) one. First, the analytical model has been used to provide a deterministic mathematical framework for the interpretation and the prediction of the scrubber's performances in terms of absorbed SO 2 molar flow and SO 2 concentration at the scrubber exit. Then, data mining and AI techniques have been applied to develop an Artificial Neural Network able to predict the error between the actual SO 2 concentration at the scrubber exit and the corresponding analytical model predictions. The final result is a combined model providing superior robustness and accuracy in the prediction of the scrubber performance while preserving a rationale for process design and operation. This interesting outcome suggests that the development of combined, or hybrid, Analytical/AI models can be a reliable and cost-effective way to improve chemical engineers' ability to design and control marine scrubbers, as well as other chemical equipment. • The performances and operation of a marine scrubber has been collected in a dataset. • An Analytical (A) model is used to describe the scrubber performances. • An Artificial Intelligence (AI) model is used to interpret the error of the A model. • The combined A/AI model greatly improve the prediction of scrubber performances. [ABSTRACT FROM AUTHOR]

Event temporal relation extraction is an important task for information extraction. The existing methods usually rely on feature engineering and require post-process to achieve optimization, though inconsistent optimization may occur in the post-process module and main neural network due to their independence. Recently, a few works start to incorporate the temporal logic rules into the neural network and achieve joint optimization. However, these methods still suffer from two shortcomings: (1) Although the joint optimization is applied, the differences between rules are neglected in the unified design of rule losses and further the interpretability and flexibility of the design of model are reduced. (2) Because of lacking abundant syntactic connections between events and rule-match features, the performance of the model may be suppressed by the inefficient interaction in training between features and rules. To tackle these issues, this paper proposes PIPER, a logic-driven deep contrastive optimization pipeline for event temporal reasoning. Specifically, we apply joint optimization (including multi-stage and single-stage joint paradigms) by combining independent rule losses (i.e., flexibility) to make PIPER more interpretable. Also, by proposing a hierarchical graph distillation network to obtain more abundant syntactic information, the designed rule-match features can effectively aid in the interaction between low-level features and high-level rules during training. The final experiments on TB-Dense and MATRES demonstrate that the proposed model can achieve competitive performance compared with the recent advances. [ABSTRACT FROM AUTHOR]

The real estate market has a significant impact on people's daily life. Therefore, it is crucial to understand the real estate market from both spatial and temporal perspectives, while there is still a lack of research in real estate industries. In this paper, a regional house price mining and forecasting (RHPMF) framework is proposed to help people intuitively understand the spatial distribution and temporal evolution of the urban estate market based on real-world housing data and urban contexts such as demographics and criminal records. Specifically, the RHPMF framework introduces a context-aware matrix factorization to extract crucial spatial and temporal price factors for revealing the housing market. Meanwhile, the RHPMF can forecast future regional house prices by manipulating the two price factors. Consequently, this study presents extensive exploratory analysis and experiments in Virginia Beach, Philadelphia, and Los Angeles to verify the proposed RHPMF. These case studies indicate that the RHPMF framework can accurately capture the market's spatial distribution and temporal evolution and forecast future regional house prices compared with recent baselines. The experimental results suggest the great potential of the proposed RHPMF for applications in real estate industries. • Unleash a new perspective to understand the regional market of real estate. • A context-aware matrix factorization is proposed to fuse information. • The decomposed factors explain the spatial and temporal evolution. • The fused information improves the forecasting accuracy of house prices. [ABSTRACT FROM AUTHOR]

Entity relation extraction aims to identify the semantic relation between entity pairs in a sentence and is an important technical support for downstream tasks such as question answering systems and semantic searching. Existing relation extraction models mainly rely on neural networks to extract the semantic information of sentences, ignoring the critical role of important phrase information in relation extraction. For this problem, this paper proposes a relation extraction model based on BERT gated multi-window attention network (BERT-GMAN). The model first uses BERT to extract the semantic representation features of the sentence and its constraint information. Secondly, it constructs the key phrases extraction network to obtain multi-granularity phrase information and uses element-wise max pooling to select key phrases features. Thirdly, it adopts classification feature perception network to further filter and globally perceive key phrase feature to form the overall features of relation classification. Finally, it combines with Softmax classifier to perform relation extraction. The experimental results on the Semeval-2010 Task 8 dataset show that the performance of the model in this paper is further improved compared with the existing methods, and the F1-score reaches 90.25%. [ABSTRACT FROM AUTHOR]

In this paper, we compared ultrasound-assisted extraction (UAE) and conventional extraction by agitation (AE) for total phenolic compounds (TPC) extraction from yerba mate leaves (Hex paraguariensis St. Hil). Water and 50% aqueous ethanol solution were used as solvent and the effect of solvent-to-solid ratio, particle size, stirring speed (AE) and ultrasound power amplitude (UAE) was examined. Two kinetic models (i.e. Peleg model and swelling / diffusion model) have been introduced to describe the kinetics of TPC extraction. The experimental data were adequately fitted by swelling/diffusion model R² > 0.98; EP < 10%; RMSE < 0,24) that considers two mechanisms coupled at different rates. The initial extraction rate increased at a higher volume of solvent (10 ml/g), smaller particle size (0.297 mm) higher ultrasound power amplitude (10056) and the use of 50% aqueous ethanol solution. Maximum equilibrium TPC concentration was observed with the synergic effect of 50% aqueous ethanol solution and the higher power amplitude (15,0 g GAO 100 g dm). This work would provide more information on the extraction mechanisms that occur in the microstructure of yerba mate leaves for the industrial production of TPC. [ABSTRACT FROM AUTHOR]

In most Internet of Things (IoT) scenes, data missing can be unavoidable when huge number of smart devices are collecting data uninterruptedly. Therefore, data imputation can be an integral part of pre-processing before data mining. It is widely known that IoT time series show strong dependencies in both spatial and temporal dimension, and the spatial relation among the devices is in non-euclidean space. However, most machine-learning-based and deep-learning-based approaches either only take temporal features into account or only catch spatial features in euclidean space. In this paper, we propose a novel network as ST-VAE (Spatio-Temporal Variational Auto-Encoder) to address the problem above. Our architecture is mainly based on Variational Auto-Encoder (VAE). Specifically, two kinds of VAE are utilized. One is for calculating the adjacent matrix of device network which is the essential input of GCN, and the other is for data imputation task based on the spatial and temporal dependencies. Experiments conducted on different real-world and public datasets demonstrate that our ST-VAE can not only populate the missing spatio-temporal data accurately but also outperforms other state-of-art approaches from the whole. [ABSTRACT FROM AUTHOR]

Membership regularized fuzzy clustering methods apply an important prior that neighboring data points should possess similar memberships according to an affinity/similarity matrix. As result, they achieve good performance in many data mining tasks. However, these clustering methods fail to take full advantage of image spatial information in their regularizations. Their performance in image segmentation problem is still not promising. In this paper, we first focus on building a novel affinity matrix to store and present the image spatial information as the prior to help membership regularized fuzzy clustering methods get excellent segmentation results. To this end, the affinity value is calculated by the fusion of pixel and region level information to present the subtle relationship of two points in an image. In addition, to reduce the impact of image noise, we use fixed cluster centers in the iteration of algorithm, thus, the updating of membership values is only guided by the prior of fused information. Experimental results over synthetic and real image datasets demonstrate that the proposed method shows better segmentation results than state-of-the-art clustering methods. • A new membership regularized fuzzy clustering is proposed for image segmentation. • The affinity matrix is based on fusion of pixel and region level information. • The model adopts a new measurement for belongingness of each pixel to each cluster. • The algorithm uses fixed cluster centers to reduce the impact of noise. • The superiorities of proposed method are verified by the comprehensive experiments. [ABSTRACT FROM AUTHOR]

Deep learning based methods have significantly boosted the study of automatic building extraction from remote sensing images. However, delineating vectorized and regular building contours like a human does remains very challenging, due to the difficulty of the methodology, the diversity of building structures, and the imperfect imaging conditions. In this paper, we propose the first end-to-end learnable building contour extraction framework, named BuildMapper, which can directly and efficiently delineate building polygons just as a human does. BuildMapper consists of two main components: 1) a contour initialization module that generates initial building contours; and 2) a contour evolution module that performs both contour vertex deformation and reduction, which removes the need for complex empirical post-processing used in existing methods. In both components, we provide new ideas, including a learnable contour initialization method to replace the empirical methods, dynamic predicted and ground truth vertex pairing for the static vertex correspondence problem, and a lightweight encoder for vertex information extraction and aggregation, which benefit a general contour-based method; and a well-designed vertex classification head for building corner vertices detection, which casts light on direct structured building contour extraction. We also built a suitable large-scale building dataset, the WHU-Mix (vector) building dataset, to benefit the study of contour-based building extraction methods. The extensive experiments conducted on the WHU-Mix (vector) dataset, the WHU dataset, and the CrowdAI dataset verified that BuildMapper can achieve a state-of-the-art performance, with a higher mask average precision (AP) and boundary AP than both segmentation-based and contour-based methods. We also confirmed that more than 60.0/50.8% of the building polygons predicted by BuildMapper in the WHU-Mix (vector) test sets I/II, 84.2% in the WHU building test set, and 68.3% in the CrowdAI test set are on par with the manual delineation level. [ABSTRACT FROM AUTHOR]

As the key part of autonomous driving (AD), High-Definition (HD) map provides more precise location and rich semantic information than the traditional map. With the development of AD, it is eager to construct HD map accurately and efficiently. Road scene perception, especially the road object detection, plays important role in HD map generation. Recently, numerous methods have been undertaken for different tasks in constructing HD maps. To stimulate further research, this paper presents a comprehensive review of recent progress in road surface's information extraction and road object detection. For road surface information extraction, it covers three main categories: road pavement, marking and line. For road object detection, it focuses on the full-element road object detection, including sign-objects, pole-like objects, guardrails, street trees and moving objects detection. In addition, the common public datasets and corresponding results are provided. Besides, for the future research directions, we also summarize the current development trends and topics. [ABSTRACT FROM AUTHOR]

The development of intelligence and informatization is the inevitable trend of safety production in coal enterprises. Data mining technology plays an important role in promoting the development of coal mine safety management. This paper uses natural language processing technology and the Bi-LSTM model to automatically classify the hidden danger information and extract the theme of the coal mine. Then, taking the risk of a coal mine gas explosion as the starting point, the gas safety situation prediction model is constructed. Finally, an accurate management and control information system based on data mining technology and intelligent technology is developed, which realizes the information analysis of a large visual screen and the intelligent information processing of the APP. The system has been applied in the Luxi coal mine, and the application effect is good, which significantly improves the safety management ability and efficiency of the coal mine. This study can provide new ideas and methods for the dual pre-control construction of coal mines, strengthen the information processing capacity of the existing pre-control system, and promote the healthy development of the coal industry. [ABSTRACT FROM AUTHOR]

• Gas turbine life management monitors and assesses turbine status and conditions. • Inlet and exit air passages impact gas turbine performance, not physical attributes. • Study gap: An incomplete gas-turbine model limits big data's predictive analysis. • Big data analytics drive optimized informed integrated health monitoring decisions for gas turbines. • Data mining extends gas turbine life by forecasting faults for the future. In addition to power generation, gas turbines (GT) are crucial components in many engineering applications, including those for aircraft, ships, and other machinery that needs a prime mover. One of the main concerns was predicting a comprehensive GT model, including intake, exit, and power-generator characteristics. Hence, this study provides a comprehensive review of GT's main components, parameters, faults, and predicting models using machine learning and data mining applications. Forty-four publications between 2014 and 2023 were chosen for this review study based on the primary keywords associated with the paper title. This study focuses solely on the GT model prediction aspects, considering the significance of field distribution, applications of data mining, research goals, benefits, constraints, and contributions to aviation, the oil and gas industry, and marine propulsion fields. Based on the examined studies, ANN, ELM, and KF applications show promise for potential GT life management. The typical limitations of GT life prediction, as observed in 93.2% of the review papers, involved the absence of a comprehensive GT model that incorporates all vibration factors and the use of substantial amounts of real data for predictive modelling. Another notable finding that underscores the knowledge gap in this area is the utilization of complete GT units to monitor unit life and facilitate life management. Apparently, the development of GT life management is still riddled with several limitations, which call for more improvement. [ABSTRACT FROM AUTHOR]

• This paper proposes a novel unsupervised feature selection called CPSLR. • CPSLR uses a specific formula to obtain the central point matrix. Construct a link graph through the central matrix and the Laplacian matrix to retain the similarity between the data. • The link graph and the data graph form a dual graph structure, which can not only preserve more complete data information but also holds on to the manifold structure of the data. • Feature selection is conducted in the latent representation space, and interconnection information among data is mined by using latent representation learning to preserve the connections among data itself. • CPSLR applies l 2,1/2-norm constraint on the feature transformation matrix to select robust and low-redundancy features. Before conducting unsupervised feature selection, it is usually assumed that these data are independent of each other. On the contrary, real data will influence each other. Therefore, traditional feature selection methods may lose information related to each other between data. This can lead to inaccurately generated pseudo-label information and may result in poor feature selection results. To find solutions to this issue, this paper proposes robust feature selection via central point link information and sparse latent representation (CPSLR). Firstly, structure a link graph by calculating the center matrix to store the distance information from the sample to the center point. If two samples have similar distances to the center point, it can be determined that they belong to the same class. Therefore, the similarity between samples is preserved, and more accurate pseudo-label information is obtained. Secondly, CPSLR uses data graph and link graph to form a dual graph structure. It can not only retain the link information between samples but also retain the manifold structures of the samples. Then, CPSLR saves the interconnection contents between samples by sparse latent representation. That is, the constraint l 2,1 -norm is exerted on the expression of latent representation, and sparse non-redundant interconnection information is preserved. And by combining central point link information with sparse latent representation makes the interconnections between data reserved more comprehensive. That is to say, the pseudo-labels obtained are more like the real labels of the classes. Finally, CPSLR constrains the feature transformation matrix by l 2,1/2 -norm constraint so as to select robust and sparse features. CPSLR uses l 2,1/2 -norm constraint to assure that the feature transformation matrix is sparse, selecting more discriminative features, thereby obtaining the feature selection that can improve its efficiency. The experiments demonstrate that the clustering result of CPSLR outperform six classical or latest compared algorithms on eight datasets. [ABSTRACT FROM AUTHOR]

Concept-cognitive learning (CCL), an emerging intelligence learning paradigm, has recently become a popular research subject in artificial intelligence and cognitive computing. A central notion of CCL is cognitive and learning things via concepts. In this process, concepts play a fundamental role when mining and fusing knowledge from data to wisdom. With the in-depth research and expansion of CCL in scopes, goals, and methodologies, some difficulties have gradually emerged, including some vague terminology, ambiguous views, and scattered research. Hence, a systematic and comprehensive review of the development process and advanced research about CCL is particularly necessary at the moment. This paper summarizes the theoretical significance, application value, and future development potential of CCL. More importantly, by synthesizing the reviewed related research, we can acquire some interesting results and answer three essential questions: (1) why examine a cognitive and learning framework based on concept? (2) what is the concept-cognitive learning? (3) how to make concept-cognitive learning? The findings of this work could act as a valuable guide for related studies in quest of a clear understanding of the closely related research issues around concept-cognitive learning. • It is the first paper that attempts to provide an in-depth analysis of the advancement of CCL. • It is a multi-view categorization of CCL in research scopes, goals, and methodologies. • It is an elucidation of the main research gaps and suggestions for the related study of CCL. • It acquires some interesting results by synthesizing the reviewed related research of CCL. [ABSTRACT FROM AUTHOR]

With the rapid development of urbanization, increasingly more data are being acquired by intelligent transportation systems (ITSs), which is of great significance for traffic flow forecasting. Efficient intelligent traffic management systems (TMSs) depend on accurately forecasting traffic flows as well as reasonably researching and judging traffic states. However, the current prediction models used in transportation forecasting tasks are traditional temporal- and spatial-dimensional prediction approaches. Especially when faced with road channelization, the numbers of lanes and lane types increase significantly, causing significant increases in the dimensionality and complexity of traffic data, and research models utilizing the cross-interaction relationships among multiple dimensions have not been considered. This has led to unsatisfactory prediction results in complex traffic environments. Therefore, it is very important to explore the exchange–correlation of complex dimensional data and the mining of hidden attributes. This paper presents a method for performing multidimensional cross-attention and spatiotemporal graph convolution. This method fully considers cross-information and constructs an attention cross-view between every pair of dimensions among the channel, time and space domains to model the cross-dimensional dependencies of traffic data. We innovatively propose a triple cross-attention and graph convolutional network (TC-GCN), which can achieve further improved traffic forecasting performance. The TC-GCN is verified on two real-world traffic datasets, namely, METR-LA and PEMS-BAY, and the experimental results are compared with those of multiple advanced baselines, showing that the proposed approach outperforms most of these baselines, which proves the effectiveness of the method proposed in this paper. • TC-GCN: A framework capturing time/space/feature dimension interrelations. • It addresses spatiotemporal information misalignment in traffic analysis. • Features tailored readout for spatiotemporal and time/feature perspectives. • We introduce time block mechanism for dynamic temporal data preservation. [ABSTRACT FROM AUTHOR]

One of the most significant sources of information from customers is customer complaints. Successful and effective complaint management can end complaint crises and ensure client loyalty, which is a sign of great service performance. In this paper, we proposed a novel customer temperament-centered and e-CCH system-based data collection and data mining method titled "3D" model for customer complaint data analysis. Three phases are (1) Development and launch of e-Customer Complaint Handling system, (2) Data collection and transfer of learning by e-Customer Complaint Handling system, and (3) Data mining by e-Customer Complaint Handling system. An advanced e lectronic C ustomer C omplaint H andling System called the e-CCH system was then developed and launched. This system adapts the seasonal associations model based on Hippocrates's customer temperament theory to the whole stages of customer complaint reporting and handling. With this system, we conducted a dataset collection work from restaurant chains of two brands over four years. As a result, we collect thousands of real-world temperament-centred customer complaint cases by four years to form the one-of-a-kind CCH dataset. This one-of-a-kind CCH dataset was open-sourced with detailed customer complaint attributes and heuristic decision-making for valuable industrial handling manner. After further analysis of this dataset, we found that customers with different temperament types tend to have different types of complaints. In addition, adapting the temperament theory to the e-CCH system can classify customer types better and provide personalized solutions. To our best knowledge, this rich and the one-of-a-kind CCH dataset reported in this paper is the first comprehensive study of customer complaint handling in an industrial service management context. Meanwhile, data mining with cross analysis and correspondence analysis and an ChatGPT experiment for transfer of learning based on this yearly and one-of-a-kind industrial customer complaint dataset was analyzed and discussed. In addition, how this dataset may contribute to more realistic complaint-handling theoretic studies for better service failure recovery and interactive marketing is discussed in-depth. [ABSTRACT FROM AUTHOR]

• A feature selection method based on binary dandelion algorithm is proposed. • The algorithm applies a sine-cosine operator and a restart strategy. • Mutual information and quick bit mutation improve the performance of the algorithm. • The algorithm performs well on datasets with larger dimensions. Feature selection (FS) is an important data preprocessing technology for machine learning and data mining. Metaheuristic algorithm (MH) has been widely used in feature selection because of its powerful search function. This paper presents an improved Binary Dandelion Algorithm using Sine Cosine operator and Restart strategy (SCRBDA) for feature selection. First, the sine cosine operator is used in the radius formula of the core dandelions (CD), which significantly enhances the ability of algorithm development and exploration. Secondly, the algorithm uses a restart strategy to increase its ability to get rid of local optimum. Thirdly, mutual information is used to guide the generation of some dandelions, which pays more attention to the correlation between the selected features and categories. Finally, quick bit mutation is used as the mutation strategy to improve the diversity of the population. The SCRBDA proposed in this paper was tested on 18 datasets of different sizes from UCI machine learning database. The SCRBDA was compared with 8 other classical feature selection algorithms, and the performance of the proposed algorithm was evaluated through feature subset size, classification accuracy, fitness value, and F1-score. The experimental results show that SCRBDA achieves the best performance, which has stronger feature reduction ability and achieves better overall performance on most datasets. Especially on large-scale datasets, SCRBDA can obtain extremely smaller feature subsets while maintaining much higher classification accuracy, and satisfactory F1-score. [ABSTRACT FROM AUTHOR]

The Internet of Things (IoT) has become an essential part of our daily lives. However, with the increasing use of IoT, the number of botnet attacks targeting resource-constrained IoT devices is also on the rise. To mitigate these threats, intrusion detection systems (IDSs) have been developed. However, traditional IDSs based on heavyweight deep/machine learning, fuzzy logic, rough set theories, or data mining techniques, often lack in detection accuracy and energy efficiency. Therefore, there is a crucial need for more lightweight, accurate, and energy-efficient IDSs capable of detecting a broad spectrum of cyber attacks. This paper presents a solution to these challenges by introducing lightweight and accurate IDSs that use a stochastic gradient descent classifier (SGDC) and four feature-selection algorithms based on a ridge regressor. The hyperparameters of the SGDC algorithm and ridge regressor model were fine-tuned to enhance the accuracy of IDSs while reducing computational complexity. Moreover, the fine-tuned feature selectors were used to decrease the dataset's dimensionality and improve the accuracy of IDSs. To evaluate the proposed IDSs, three network traffic datasets (KDD-CUP-1999, BotIoT-2018, and N-BaIoT-2021) were used. The systems achieved an average accuracy of 92.69%, and the number of features was reduced by an average of 79.93%. The results demonstrate that the proposed systems can be utilized for lightweight IDSs on resource-constrained IoT devices. Overall, this paper presents a significant contribution to the field of IDSs for IoT devices by offering an efficient and accurate solution. The proposed lightweight IDSs have the potential to enhance IoT security and privacy, safeguarding sensitive IoT data. • SGDC-based lightweight IDSs were developed for resource-constrained IoT devices. • The optimal hyperparameters of SGDC were determined with a grid-search algorithm. • The fine-tuned SGDC boosted accuracy and reduced IDSs packet processing time greatly. • The ridge regression-based feature selectors determined the most efficient features. • The fine-tuned feature selectors reduced feature space greatly and improved accuracy. [ABSTRACT FROM AUTHOR]

Sorting is a fundamental component in many different applications including web indexing engines, geographic information systems, data mining and database systems. However, the existing optimal algorithms still cause huge costs when the input data becomes very large. Thus, the approximate sorting for big data is considered in this paper. The goal of approximate sorting for big data is to generate an approximate sorted result, but using less CPU and I/O cost. For big data, we consider the approximate sorting and applications for approximate sorted results in I/O model. The quality of approximate sorting results is usually measured by the distance metrics on permutation space. However, the existing metrics on permutation space are not available for external approximate sorting algorithms. Thus, we propose a new kind of metric named External metric, which ignores the errors and dislocations that happened in each I/O block. The External metric of Spearman's footrule metric is named as External Spearman's footrule metric (short as ESP metric), which is analyzed in this paper. Furthermore, to facilitate a better evaluation of the approximate sorted result, we propose a new metric, named as errors , which directly states the number of dislocation of the elements. The External errors are also considered in this paper. Then, according to the rate-distortion relationship under these two metrics, the lower bound of these two metrics on external approximate sorting problem with t I/O operations is proved. We propose a k-passes external approximate sorting algorithm, named as EASORT, and prove that EASORT is asymptotically optimal under ESP metric. Finally, we consider the applications on approximate sorting results. An index for the approximate sorted result is proposed. The single and range query on approximate sorted result using this index are also analyzed. Further, the sort-merge join on two relations, where one of the relations is approximate sorted or both relations are approximate sorted, are all discussed in this paper. • This paper first discusses the approximate sorting in I/O model. • This paper proposes a new external metric to evaluate the approximate sorted result. • This paper analyzes the lower bound of two metrics of approximate sorted result. • This paper proposes a new external approximate sorting algorithm. • This paper analyzes the applications on approximate sorted result. [ABSTRACT FROM AUTHOR]

In an aero-engine rotor system, a blade-casing rub-impact fault can lead to a frequency modulation phenomenon of the vibration signal. This paper discusses the extraction of the modulation information in the fault signal using the variational nonlinear chirp mode decomposition (VNCMD) method to detect early rub-impact faults. The regularities of frequency modulation are revealed in both the simulated and experimental blade-casing rubbing signals. According to the frequency modulation law, an empirical formula is summarized to judge the rub-impact fault type. Then, we propose an evaluation index to judge the fault severity based on the energy theory by using the advantage of the VNCMD method that can reconstruct the decomposed fault signal. Finally, the fault severity in the simulated and experimental signals with different parameters is compared. The results show the correctness and reliability of this evaluation index. The research in this paper provides a theoretical basis for the detection and judgment of the rubbing severity of blade-casing rub-impact faults. • The time-frequency modulation rules of different rub-impact faults are found and summarized. • An empirical formula used to judge the type of rubbing is summarized in the early stage of the rub-impact fault. • An evaluation index for judging the fault severity is proposed based on energy theory. [ABSTRACT FROM AUTHOR]

• An extensive literature review was conducted with keywords of "data collection", "information extraction", and "autonomous vehicles". • Over 160 studies has been screened and 27 studies have been selected for detailed analysis. • Selected studies focused on driving behavior modeling, with aspects of data collection and information extraction. • Findings highlighted the methods and equipment used for data collection, features collected, size and frequency of the data, and associated main challenges. • A data-analytics framework was developed with following components: data collection, data processing, data storage, data mining and analytics, data sharing, along with external legal and ethical considerations. • A data fusion framework was proposed with two layers of analytics, and a pathway to transfer the knowledge across different transport sectors. When talking about automation, "autonomous vehicles", often abbreviated as AVs, come to mind. In transitioning from the "driver" mode to the different automation levels, there is an inevitable need for modeling driving behavior. This often happens through data collection from experiments and studies, but also information extraction, a key step in behavioral modeling. Particularly, naturalistic driving studies and field operational trials are used to collect meaningful data on drivers' interactions in real–world conditions. On the other hand, information extraction methods allow to predict or mimic driving behavior, by using a set of statistical learning methods. In simple words, the way to understand drivers' needs and wants in the era of automation can be represented in a data–information cycle, starting from data collection, and ending with information extraction. To develop this cycle, this research reviews studies with keywords "data collection", "information extraction", "AVs", while keeping the focus on driving behavior. The resulting review led to a screening of about 161 papers, out of which about 30 were selected for a detailed analysis. The analysis included an investigation of the methods and equipment used for data collection, the features collected, the size and frequency of the data along with the main problems associated with the different sensory equipment; the studies also looked at the models used to extract information, including various statistical techniques used in AV studies. This paved the way to the development of a framework for data analytics and fusion, allowing the use of highly heterogeneous data to reach the defined objectives; for this paper, the example of impacts of AVs on a network level and AV acceptance is given. The authors suggest that such a framework could be extended and transferred across the various transportation sectors. [ABSTRACT FROM AUTHOR]

Signed network embedding is an approach to learning low-dimensional representations of nodes in signed networks with both positive and negative links, which facilitates downstream tasks such as link prediction with general data mining frameworks. Due to the distinct properties and significant added value of negative links, existing signed network embedding methods usually design dedicated methods based on social theories such as balance theory and status theory. However, existing signed network embedding methods ignore the characteristics of multiple facets of each node and mix them up in one single representation, which limits the ability to capture the fine-grained attentions between node pairs. In this paper, we propose MUSE , a MU lti-faceted attention-based S igned network E mbedding framework to tackle this problem. Specifically, a joint intra- and inter-facet attention mechanism is introduced to aggregate fine-grained information from neighbor nodes. Moreover, balance theory is also utilized to guide information aggregation from multi-order balanced and unbalanced neighbors. Experimental results on four real-world signed network datasets demonstrate the effectiveness of our proposed framework. [ABSTRACT FROM AUTHOR]

Continual learning is an emerging research branch of deep learning, which aims to learn a model for a series of tasks continually without forgetting knowledge obtained from previous tasks. Despite receiving a lot of attention in the research community, temporal-based continual learning techniques are still underutilized. In this paper, we address the problem of temporal-based continual learning by allowing a model to continuously learn on temporal data. To solve the catastrophic forgetting problem of learning temporal data in task incremental scenarios , in this research, we propose a novel method based on attentive recurrent neural networks , called Temporal Teacher Distillation (TTD). TTD solves the catastrophic forgetting problem in an attentive recurrent neural network based on three hypotheses, namely Rotation Hypothesis, Redundant Hypothesis, and Recover Hypothesis. Rotation Hypothesis and Redundant hypotheses could cause the attention shift phenomenon, which degrades the model performance on the learned tasks. Moreover, not considering the Recover Hypothesis increases extra memory usage in continuously training different tasks. Therefore, the proposed TTD based on the above hypotheses complements the inadequacy of the existing methods for temporal-based continual learning. For evaluating the performance of our proposed method in task incremental setting, we use a public dataset, WIreless Sensor Data Mining (WISDM) , and a synthetic dataset, Split-QuickDraw-100. According to experimental results, the proposed TTD significantly outperforms state-of-the-art methods by up to 14.6% and 45.1% in terms of accuracy and forgetting measures, respectively. To the best of our knowledge, this is the first work that studies continual learning in real-world incremental categories for temporal data classification with attentive recurrent neural networks and provides the proper application-oriented scenario. [ABSTRACT FROM AUTHOR]

The goal of this paper is to improve the predictive accuracy of data streaming algorithms without increasing the processing time of the incoming data. We propose the EnHAT (Ensemble Combined with Hoeffding Adaptive Tree) algorithm, which combines the state-of-the-art Hoeffding Adaptive Tree (HAT) algorithm with an ensemble of J48 decision trees induced from sequential chunks of the data stream. The slack time of HAT adaptation to a new window of incoming records is utilized in parallel for building a decision-tree ensemble. In our experiments on 4 benchmark streaming datasets and 4 synthetic datasets with different types of concept drift, EnHAT has reached the highest predictive accuracy in 23 out of 26 cases compared to an ensemble of J48 trees, HAT, and a single J48 model induced from the last sliding window. Thus, we can conclude that the ensemble/HAT synergy yields better prediction results than each one of the two approaches on its own. The higher accuracy does not come at the expense of any additional computational effort beyond the model induction times of the combined algorithms. • We build interpretable and accurate classification models from massive data streams. • The Hoeffding Adaptive Tree is combined with an ensemble of J48 decision trees. • The synergy between the two algorithms improves classification accuracy. • The proposed algorithm does not increase the model induction time. [ABSTRACT FROM AUTHOR]

Compared with relatively easy feature creation or generation in data analysis, manual data labeling needs a lot of time and effort in most cases. Even if automated data labeling​ seems to make it better in some cases, the labeling results still need to be checked and verified by manual. The High Dimension and Low Sample Size (HDLSS) data are therefore very common in data mining and machine learning. For classification problems with the HDLSS data, due to data piling and approximate equidistance between any two input points in high-dimension space, some traditional classifiers often give poor predictive performance. In this paper, we propose a Maximum Decentral Projection Margin Classifier (MDPMC) in the framework of a Support Vector Classifier (SVC). In the MDPMC model, the constraints of maximizing the projection distance between decentralized input points and their supporting hyperplane are integrated into the SVC model in addition to maximizing the margin of two supporting hyperplanes. On ten real HDLSS datasets, the experiment results show that the proposed MDPMC approach can deal well with data piling and approximate equidistance problems. Compared with SVC with Linear Kernel (SVC-LK) and Radial Basis Function Kernel (SVC-RBFK), Distance Weighted Discrimination (DWD), weighted DWD (wDWD), Distance-Weighted Support Vector Machine (DWSVM), Population-Guided Large Margin Classifier (PGLMC), and Data Maximum Dispersion Classifier (DMDC), MDPMC obtains better predictive accuracy and lower classification errors than the other seven classifiers on the HDLSS data. • A novel MDPMC approach is proposed for HDLSS problems. • Maximum decentral projection is added to the constraints of MDPMC. • Data piling in the HDLSS setting can be solved by the MDPMC model. • The MDPMC model deals well with the approximate equidistance in HDLSS data. • Our proposed MDPMC outperforms the other seven classifiers for predictive performance. [ABSTRACT FROM AUTHOR]

• This study produces predictive models for MaaS use by applying artificial neural network (ANN) analysis. • MaaS use models are developed separately for social, general, and work trips. • The cross industry standard process and data mining framework is followed in designing ANN-based predictive models. • Different sets and combinations of socio-economic, personal travel and Covid-19 related factors are key predictors of MaaS use for social, general, and work trips. • Findings can inform strategies and policies to promote MaaS use. Mobility as a Service (MaaS) has gained popularity as a means of sustainable urban transport which makes the understanding of MaaS use critical for its design and promotion. Studies have contributed to the growing understanding of MaaS, its design features, and consumers' willingness to pay for MaaS in general, without recourse to the differences in traveling for work, social and general purposes. This paper develops and tests models to predict MaaS use by applying artificial neural network analysis, following the cross industry standard process and data mining framework. It also estimates separate models for social, general, and work trips, using 33 input variables reflecting network externality, transaction cost, behaviour, institutional, environmental concern, personal travel, and socio-economic factors. Data are collected from a survey of 331 Australians. The study reveals different sets of socio-economic factors, the impact of Covid-19, and personal travel factors as key predictors of MaaS use for general, social, and work trips with an average prediction accuracy of 68%, 68%, and 75% respectively. The findings can be used to inform strategies and policies on how to attract a user base with respect to socio-economic and personal travel factors for promoting MaaS use. [ABSTRACT FROM AUTHOR]

Dynamic networks are popularly used to describe networks that change with time. Although there have been a large number of research works on understanding dynamic networks using link prediction, node classification and community detection, there is rare work that is specially designed to address the challenge of big network size of dynamic networks. To this end, we study in this paper an emerging and challenging problem of network coarsening in dynamic networks. Network coarsening refers to a class of network "zoom-out" operations where node pairs and edges are grouped together for efficient analysis on big networks. However, existing network coarsening approaches can only handle static networks where network structure weights have been predefined before the coarsening calculation. Under the observation that big networks are highly dynamic and naturally change over time, we consider in this paper to embed information diffusion data which reflect the dynamics of networks for network coarsening. Specifically, we present a new Semi-NetCoarsen approach that jointly maximizes the likelihood of observing the information diffusion data and minimizes the network regularization with respect to the predefined network structural data. The learning function is convex and we use the accelerated proximal gradient algorithm to obtain the global optimal solution. We conduct experiments on two synthetic and five real-world data sets to validate the performance of the proposed method. [ABSTRACT FROM AUTHOR]

Ab initio calculation and data mining are very useful methods to predict the interrelationships among structure properties. In this paper, the first-principles modelling based on the density functional theory (DFT) was used to explore the structural, mechanical, and electronic properties of ten hydrides REH 2 (RE = La, Ce, Pr, Nd, Sm, Tb, Dy, Ho, Er, Tm). From the present work, it can be noted that LaH 2 , CeH 2 , NdH 2, and PrH 2 compounds crystallizing in the hydrogen-rich cubic phase present good mechanical properties which could be of use to future applications in hydrogen storage. Furthermore, the electronic structures were also evaluated to provide a deep insight regarding their trends and confirm the metallic character of these systems. While, HoH 2 , DyH 2 , ErH 2 , TbH 2 , TmH 2 , SmH 2 hydrides are mainly hard and rigid systems with high values of the bulk modulus (B), shear modulus (G), Young's modulus (E), micro-hardness (H), and low values of B / G ratio. More significantly, the adopted principal component analysis (PCA) provided an advantageous technique in selecting and identifying the minimum of inputs variables necessary to capture all information of the systems. • A DFT investigation of mechanical and electronic properties for ten rare-earth hydrides has been performed. • Elastic and mechanical properties for some dihydrides are studied. • The paper explores results connected with a new approach based on data mining. • LaH 2 , CeH 2 , NdH 2 and PrH 2 are very promising compounds for possible applications in hydrogen storage. [ABSTRACT FROM AUTHOR]

Autonomous-rail Rapid Transit (ART) is a new type of road public transportation with an automatic driving function, and it is one of the main places for epidemic spread. Scientific wearing of masks has been proven to be an extremely effective way to prevent an epidemic. However, drivers wearing masks make it difficult for traditional driver alert detection methods based on facial information to operate effectively. This paper proposes an ART driver alert detection system based on multimodal information fusion and deep learning, which can be applied to drivers wearing masks. The detection system is mainly divided into three modules: data acquisition module, preprocessing module, and vigilance detection module. The data acquisition module collects real-time face thermal imaging data, and environment temperature information of 13 people in the vigilance and fatigue state of the ART simulated driving environment. The preprocessing module proposes a face thermal imaging key information extraction method based on the constraints between Yolov5 and thermal image interframe constraints and extracts the mean temperature information and the hot voxel information (HV) of the eyes, forehead, and mask areas. Then the extracted information and the environment temperature information are transmitted together to the vigilance detection module for learning and classification. The vigilance detection module proposes a lightweight hybrid-attention optimization network, and this paper extracts training sets and testing sets from the data of 13 subjects. And the training set data is enhanced. Each sample is 90 frames. The maximum accuracy rate of the testing set reaches 99.57%. The feasibility of the detection system proposed in this paper is proved, which makes the thermal imaging and attention model show great potential in the application of traffic safety operation and maintenance. • A contactless vigilance detection method is proposed for ART drivers with face masks. • The multi-modal fusion method uses facial thermal image and environmental data. • A lightweight network model that integrates hybrid-attention mechanism is proposed. • A method for extracting face key information from thermal images is proposed. • The testing set shows that the system is feasible and the accuracy rate is 99.57%. [ABSTRACT FROM AUTHOR]