Search

In order to better achieve the integration and innovation of sports social platforms and information technology, the author proposes an SFD (Sport Friend Discover) sports friend recommendation model based on physical testing big data. The core idea of this model is to use physical measurement data to match the similarity between athletes and recommend suitable exercise partners. Specifically, we collected a large amount of physical measurement data, including height, weight, body fat percentage, muscle mass, etc. Then, through data mining algorithms, these data are transformed into feature vectors of the movers. Next, we use a similarity algorithm to calculate the similarity between different athletes and find the most matching motion partner with the user. The results show that the SFD method outperforms the other two traditional recommendation methods on the dataset, with P @ 10, P @ 20, P @ 30, and P @ 40 of SFD reaching 0.099, 0.095, 0.085, and 0.591, respectively. SFD not only utilizes more neighboring information than FOAF based on local graph structure, but also compared to TRW based on global graph structure method, at the same time, the importance of the node itself is also considered, resulting in higher accuracy. It has been proven that the SFD sports friend recommendation model based on physical testing big data has achieved good results in recommending sports partners. Users can quickly find sports partners with similar body types and health conditions, improving the fun and effectiveness of exercise. [ABSTRACT FROM AUTHOR]

The extraction of webpage information is of paramount importance in the realm of competitive intelligence. This research is dedicated to the design and implementation of a visual webpage information extraction module within a competitive intelligence system, approached through the lens of research and development (R&D) technology and its practical applications. Initially, the study delineates the objectives and requirements for webpage information extraction, emphasizing the practical needs of competitive intelligence systems. By critically assessing the strengths and weaknesses of current theories and methodologies in webpage text information extraction, this paper introduces an innovative visual method for extracting webpage text information. Subsequently, the paper meticulously outlines the comprehensive architecture of the proposed module. Building upon this foundation, the study delves into the specifics of the extraction template, rule generation, optimization techniques, and the extraction algorithm pivotal to the process of visual webpage information extraction. The system's effectiveness and practical utility are substantiated through a series of confirmatory experiments, the results of which are thoroughly analyzed. The findings affirm that the developed system adeptly fulfills the webpage information extraction needs of competitive intelligence systems, contributing significantly to the R&D efforts in which the authors are engaged. [ABSTRACT FROM AUTHOR]

Recently hybrid cloud solutions integrating public and private cloud is proposed to address the privacy and security concerns faced by Enterprises in their data offloading decisions. In these solutions, the transformed data is kept in public cloud while transformation keys are kept in private cloud. The existing works for data transformation used in hybrid clouds does not address multiple objectives of privacy, security, fine grained access control, utility preservation for mining and data retrieval efficiency. This work proposes a multi objective data transformation technique for hybrid cloud to address all these objectives. The proposed solution is built on attribute based hierarchical data access control with hierarchy selection based on joint consideration of security, utility preservation and retrieval efficiency. The proposed solution is able to provide 5% higher security strength, 1.34% higher clustering accuracy over perturbed data and 29.95 % higher data retrieval efficiency over perturbed data compared to existing works. [ABSTRACT FROM AUTHOR]

With the proliferation of digital platforms, the dissemination of literary works has encountered unprecedented challenges, particularly concerning copyright infringement and unauthorized use. This study introduces a comprehensive framework for copyright protection and risk assessment, specifically tailored to online literary works. The framework employs advanced CNN based information extraction (IE) techniques coupled with machine learning (ML) algorithms to identify, classify, and protect literary content against copyright violations. Firstly, we delineate a novel CNN-Decision tree-based IE methodology that systematically harvests metadata and textual content from various online repositories. This process is designed to detect and index online literary works, extracting pertinent features such as authorship, publication date, and textual patterns. Following the extraction, the study utilizes natural language processing (NLP) to analyze and compare content, pinpointing potential instances of copyright infringement by identifying significant overlaps and stylistic similarities with registered works. Subsequently, we introduce a risk assessment model developed through supervised machine learning. This model is trained on a labelled dataset comprising instances of both copyrighted and non-copyrighted works, along with known cases of copyright infringement. By analyzing the extracted features, the model assesses the probability of infringement, categorizing risks into high, medium, and low categories. This stratification allows stakeholders to prioritize enforcement actions and resources efficiently. The study further explores the implementation of various ML algorithms, including decision trees, support vector machines, and neural networks, to determine the most effective approach for copyright protection in the literary domain. We evaluate the models based on accuracy, precision, recall, and F1-score metrics, emphasizing their capacity to generalize and operate in dynamic, real-world environments. [ABSTRACT FROM AUTHOR]

With the development of network technology, online shopping is becoming more and more convenient. But the increasing number of products also makes it difficult for consumers to make the right decision. When there is no apparent market demand, how to recommend products with commercial potential to customers has become an urgent problem for businesses to solve. This paper proposes e-commerce product recommendation based on user preference and association rule algorithm aiming at the problems existing in e-commerce product recommendation. Firstly, this paper constructs a user interest modeling method. Through analyzing users' interests and preferences, to provide users with timely and accurate personalized services. Then, the FP_Growth algorithm is optimized and improved. A more effective CTE-MARM algorithm is designed, and an association rules database based on user benefit items is constructed and analyzed jointly. Analyze products with strong correlations. According to consumers' interest levels, TOP-N is the best product choice. Experiments show that the algorithm has higher prediction accuracy. The research results of this project can not only improve enterprises' ability to analyse data and provide data support for enterprises to carry out effective marketing management. [ABSTRACT FROM AUTHOR]

This paper uses the hybrid leapfrog algorithm to mine user information in encrypted advertisements effectively and intelligently. This method handles the nonlinearity of the original data by mapping it into kernel space. The representation of the original ciphertext in the kernel space is obtained by sparsely reconstructing the encrypted original advertising data. Build a corresponding scoring mechanism and select the best advertising data characteristics. The selected data were clustered using the data fuzzy clustering method based on the improved hybrid leapfrog. Set the adjustment coefficient to improve the local optimization performance of hybrid frog leaping. This algorithm uses the tightness and separation in genetic algorithms and constructs a fitness function to determine the clustering critical value. This enables the practical, intelligent mining of homomorphic passwords with privacy protection. Experimental results show that the method proposed in this article can effectively improve the convergence speed and accuracy of clustering. Improve Blowfish by combining multi-threading, sharing encryption and other methods. This enables encryption and decryption of large amounts of model data. The research of this project has very important research value in improving the security performance and effectiveness of cryptographic algorithms. [ABSTRACT FROM AUTHOR]

The main objective of the study is to address the lack of comprehensive management technology in student campus life in universities. Starting from the life trajectory data of students in campus life, a trajectory mining model combining data mining technology and university information system is designed. In addition, an applied clustering algorithm is designed to classify different trajectory feature types. The research results show that in actual trajectory analysis, the categories of action trajectories from dormitories to canteens, from 0 to 4, are 87.64%, 87.86%, 86.97%, 88.63%, and 88.71%, respectively, which are the most matched effective action trajectories. It can be seen that the trajectory analysis model designed in the study is effective and can provide assistance for the comprehensive academic management of college students. [ABSTRACT FROM AUTHOR]

Assessing tourists' social media perceptions is crucial in today's digital age. Due to its industrial reputation, Johor Bahru, Malaysia, has been overlooked for heritage tourism. This data mining study examines tourists' experiences and motivations for visiting Johor Bahru's heritage buildings of 2,369 Google Review and Tripadvisor comments using web crawlers scripted in Python. The content analysis identified six themes associated with tourist motivation: activity, accessibility, diversity, attachment, inclusivity, and appearance. Even though their aesthetics are subdued, Johor Bahru's public historic sites thrive on accessibility and active activities, unlike magnificent royal edifices. Despite their architectural value, royal structures attracted fewer tourists due to the absence of diversity, inclusivity, and activities and restrictive measures. Maintaining the city's legacy and architectural identity requires highlighting these buildings. This study accentuates social media-driven popular views. The findings may encourage local authorities and stakeholders to strengthen Johor Bahru's identity and promote heritage tourism. [ABSTRACT FROM AUTHOR]

Copyright of Baghdad Science Journal is the property of Republic of Iraq Ministry of Higher Education & Scientific Research (MOHESR) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Clinical named entity extraction (NER) based on deep learning gained much attention among researchers and data analysts. This paper proposes a NER approach to extract valuable Parkinson’s disease-related information. To develop an effective NER method and to handle problems in disease data analytics, a unique NER technique applies a “recognize-map-extract (RME)” mechanism and aims to deal with complex relationships present in the data. Due to the fast-growing medical data, there is a challenge in the development of suitable deep-learning methods for NER. Furthermore, the traditional machine learning approaches rely on the time-consuming process of creating corpora and cannot extract information for specific needs and locations in certain situations. This paper presents a clinical NER approach based on a convolutional neural network (CNN) for better use of specific features around medical entities and analyzes the performance of the proposed approach through fine-tuning NER with effective pre-training on the BC5CDR dataset. The proposed method uses annotation of entities for various medical concepts. The second stage develops a clinically NER method. This proposed method shows interesting results on the performance measures achieving a precision of 92.57%, recall of 92.22%, and F1- measure of 91.6% [ABSTRACT FROM AUTHOR]

In recent years, tensor problems have been studied in multiple fields of science and engineering including applied mathematics, the theory of completely integrable quantum, data mining, statistics, physics, chemistry, machine learning, medical engineering, and others. In machine learning, the word tensor informally refers to two different concepts that organize and represent data. In this work, at first, the concept of the sign function of a tensor is developed using the sign function of a matrix. Then, we propose an iterative method to find the sign function of a tensor. We prove that the order of convergence of the proposed method is three. Finally, we extend the iterative method for solving the Young–Baxter equation, which has many applications in fully integrable quantum theory, classical systems, and exactly solvable models of statistical physics. The accuracy and effectiveness of the proposed method in comparison to well-known methods are demonstrated by various numerical examples. [ABSTRACT FROM AUTHOR]

Transportation infrastructures and operations can be severely impacted during flood events, leading to significant disruptions to the flow of goods and services. Although numerous studies have evaluated the direct impacts of flood events on the performance of transportation infrastructures, the indirect impacts or cascading effects have been rarely assessed. Hence, this paper examines the cascading effects of floods on transportation infrastructure using data mining algorithms. First, 33 effects of flood events on transportation infrastructure have been identified based on data collected for multiple flood events in New York and New Jersey. Second, association rule mining analysis was implemented to identify the key co-occurrences between flooding and the different events. Third, network analysis was conducted to quantify the co-occurrences or key combinations among the events. Fourth, cluster analysis was used to group or prioritize the cascading effects and co-occurring events into highly connected clusters to identify the most critical ones based on two scenarios: (1) without consideration of co-occurrences (Scenario 1); and (2) with consideration of co-occurrences (Scenario 2). The findings provided insights that while some cascading impacts could be individually critical/frequent (under Scenario 1), other cascading impacts could also result due to a combination of different effects that might not be perceived to be critical on the individual level but rather become critical when combined with other cascading events (under Scenario 2). The outcomes of this paper demonstrate the importance of considering the co-occurrences between the events and cascading effects, rather than analyzing them in isolation. This study adds to the body of knowledge by offering an analytical approach that could be used to identify and prioritize critical cascading effects of flood events on the operations and performance of transportation infrastructures. [ABSTRACT FROM AUTHOR]

The aim of this research is to develop a robust multicriteria method to classify suppliers into ordered categories and its validation in real contexts. The proposed technique is based on a property of net flows of the PROMETHEE method and uses global and local search concepts, which are common in the optimisation field. The results obtained are compared to those from the most cited sorting algorithm, and an empirical validation and sensitivity analysis is performed using real supplier evaluation data. Furthermore, it does not require additional information from decision‐makers as other sorting algorithms do for assigning incomparable or indifferent alternatives to groups. An extension of the silhouette concept from data mining is also contributed to measure the quality of ordered classes. Both contributions are easy to apply and integrate into decision support systems for automated decisions in the supply chain management. Finally, this practical approach is also useful to classify customers and any type of alternatives or actions into ordered categories, which have an increasing number of real applications. [ABSTRACT FROM AUTHOR]

Educational big data has great predictive power to identify academically at-risk students. Based on fully mining educational big data, this study developed an early alert dynamic model of academically at-risk students to confirm and extend this proposition. An empirical test was carried out to validate the reliability and validity of the dynamic model by using data analysis methods. The results proved that the developed model could precisely identify academically at-risk students, scientifically assess their academic risk status, and dynamically provide personalised interventions. The overall prediction success rate of the developed model was 80.8%. In addition, application reflections were discussed to shed light on future implementation. [ABSTRACT FROM AUTHOR]

Despite recent improvements in collected drilling data quality and volume, the actual number of wells being used in studies remain low and are often limited to a single source and oil field, producing results that are prone to overfitting and are non-transferable.In our study, we access oil drilling data from 5 of more than 20 oil drilling companies collected from 2005 to 2016 from our industrial partner to create well drilling duration models for well planning. This project could lead to the creation of more generalized models from larger datasets than others in literature. However, the data is difficult to process without expert knowledge, further complicated by properties such as unharmonized, source-locked, semantic heterogeneity, sparse and unlabelled. Conventional automated methods for feature selection, propositionalization, multi-source, or block-wise missing techniques could not be used.In this paper, we describe our method to assist the Knowledge Discovery in Databases (KDD) Selection stage of the abovementioned data - Feature Selection before Propositionalization (FSbP) via Database Attribute Health Feature Reduction (DAHFR) and Report Feature Correlation Matrix (RFCM), collectively known as FvDR. DAHFR and RFCM are filter-type feature selection techniques that could measure relational missingness and keyword correlations respectively despite the complexity of multi-source oil drilling data. FvDR successfully reduced the scope from 700 tables containing 20,000 columns to 22 tables containing fewer than 707 columns while successfully selecting 13 of 16 relevant tables suggested by literature. Despite the loss of information from limitations of subsequent KDD procedures, preliminary models show promising results with over half the test predictions falling within the 20% error margin required for well planning. FvDR proves to be indispensable in KDD as a FSbP framework as it reduces features for examination and streamlines the research process necessary to understand business rules for data harmonization and propositionalization. [ABSTRACT FROM AUTHOR]

Rapidly emerging generative artificial intelligence (GenAI) models stand at the epicentre of current public discourse. They demonstrate impressive abilities to generate various types of data promptly and cost-effectively. However, AI developers need to train their systems on massive volumes of data which is usually copyrighted. Therefore, the growth of copyright-related concerns in the field of GenAI comes as no surprise. The study introduces two solutions which could mitigate the tension between copyright holders and AI developers, one legal (text and data mining (TDM) exceptions of the CDSM Directive) and one technical (synthetic data), highlighting the promises and challenges of both. First, the article will discuss the capability of TDM exceptions to facilitate the fundamental right to information and the freedom of research in the context of AI development. Next, the paper will analyse how providers of GenAI models can leverage synthetic data to comply with copyright law while training their systems and what risks might be associated with this approach. The findings of this study will indicate what issues, in both legal and technical spheres, should be addressed to ensure a balance of powers in the digital environment and effective functionality of the EU AI sector. [ABSTRACT FROM AUTHOR]

Background: Dementia is a leading factor in the institutionalization of older adults. Informal caregivers' desire to institutionalize (DI) their care recipient with dementia (PwD) is a primary predictor of institutionalization. This study aims to develop a prediction model for caregivers' DI by mining data from an eHealth platform in a high-prevalence dementia country. Methods: Cross-sectional data were collected from caregivers registering on isupport-portugal.pt. One hundred and four caregivers completed the Desire to Institutionalize Scale (DIS) and were grouped into DI (DIS score ≥ 1) and no DI (DIS score = 0). Participants completed a comprehensive set of sociodemographic, clinical, and psychosocial measures, pertaining to the caregiver and the PwD, which were accounted as model predictors. The selected model was a classification tree, enabling the visualization of rules for predictions. Results: Caregivers, mostly female (82.5%), offspring of the PwD (70.2), employed (65.4%), and highly educated (M 15 years of schooling), provided intensive care (Mdn 24 h. week) over a median course of 2.8 years. Two-thirds (66.3%) endorsed at least one item on the DIS (DI group). The model, with caregivers' perceived stress as the root of the classification tree (split at 28.5 points on the Zarit Burden Interview) and including the ages of caregivers and PwD (split at 46 and 88 years, respectively), as well as cohabitation, employed five rules to predict DI. Caregivers scoring 28.5 and above on burden and caring for PwD under 88 are more prone to DI than those caring for older PwD (rules 1–2), suggesting the influence of expectations on caregiving duration. The model demonstrated high accuracy (0.83, 95%CI 0.75, 0.89), sensitivity (0.88, 95%CI 0.81, 0.95), and good specificity (0.71, 95%CI 0.56, 0.86). Conclusions: This study distilled a comprehensive range of modifiable and non-modifiable variables into a simplified, interpretable, and accurate model, particularly useful at identifying caregivers with actual DI. Considering the nature of variables within the prediction rules, this model holds promise for application to other existing datasets and as a proxy for actual institutionalization. Predicting the institutional placement of PwD is crucial for intervening on modifiable factors as caregiver burden, and for care planning and financing. [ABSTRACT FROM AUTHOR]

Purpose: Sacubitril/valsartan is extensively used in heart failure; however, there are few long-term safety studies of it in a wide range of populations. The aim of this study was to evaluate sacubitril/valsartan-induced adverse events (AEs) through data mining of the U.S. Food and Drug Administration Adverse Event Reporting System (FAERS). Methods: Reports in the FAERS from the third quarter of 2015 (FDA approval of sacubitril/valsartan) to the fourth quarter of 2023 were collected and analyzed. Disproportionality analyses, including the reporting odds ratio (ROR), the proportional reporting ratio (PRR), the Bayesian confidence propagation neural network (BCPNN), and empirical Bayesian geometric mean (EBGM) algorithms were adopted in data mining to quantify signals of sacubitril/valsartanassociated AEs. Results: A total of 12,001,275 reports of sacubitril/valsartan as the "primary suspected (PS)" and 99,651 AEs induced by sacubitril/valsartan were identified. More males than females reported AEs (59.95% vs. 33.31%), with the highest number of reports in the 60-70 years age group (8.11%), and most AEs occurred < 7 days (14.13%) and = 60 days (10.69%) after dosing. Sacubitril/valsartan-induced AE occurrence targeted 24 system organ classes (SOCs) and 294 preferred terms (PTs). Of these, 4 SOCs were strongly positive for all four algorithms, including cardiac disorders, vascular disorders, ear and labyrinth disorders, and respiratory, thoracic and mediastinal disorders. Among all PTs, consistent with the specification, hypotension (n = 10,078) had the highest number of reports, and dizziness, cough, peripheral swelling, blood potassium increased, and renal impairment were also reported in high numbers. Notably, this study also discovered a high frequency of side effects such as death, dyspnea, weight change, feeling abnormal, hearing loss, memory impairment, throat clearing, and diabetes mellitus. Conclusion: This study identified potential new AE signals and gained a more general understanding of the safety of sacubitril/valsartan, promoting its rational adoption in the cardiovascular system. [ABSTRACT FROM AUTHOR]

A tiebreak in tennis is one of the critical moments where players are expected to excel under mental pressure and maintain high level of performance. Despite the importance of tiebreak points, research exploring the performance of male and female players during such match phrase remains limited. This study aimed to investigate i) the overall tiebreak performance of male and female players in relation to the outcome, ii) to examine their point-level performance by considering different contextual variables. A total of 535 tiebreaks comprising 6380 points from the 2016–2021 US Open men’s and women’s singles matches were collected. The difference in match performance between winning and losing players within the entire tiebreak game was explored. A subsequent decision tree analysis was then used to analyse the effect of the contextual and performance variables on tiebreak point-by-point outcome. The results showed that male and female Winning players outperformed the Losing players in 1st Serve, Serve Width and Net approach performance. The analysis of point-level performance showed that Net point, Score scene, and Point server substantially impacted tennis players’ tiebreak outcome. These findings provide valuable insight for coaches and players, informing tiebreak tactics tailoring and training in relevance to different match status. [ABSTRACT FROM AUTHOR]

Education is essential for individuals to lead fulfilling lives and attain greatness by enhancing their value. It improves self-assurance and enables individuals to navigate the complexities of modern society effectively. Despite the obstacles it faces, education continues to develop. The objective of numerous pedagogical approaches is to enhance academic performance. The development of technology, especially artificial intelligence, has caused a significant change in learning. This has made instructional materials available anytime and wherever easily accessible. Higher education institutions are adding technology to conventional teaching strategies to improve learning. This work presents an innovative approach to student performance prediction in educational settings. The strategy combines the DistilBERT with LSTM (DBTM) hybrid approach with the Spotted Hyena Optimizer (SHO) to change parameters. Regarding accuracy, log loss, and execution time, the model significantly improved over earlier models. The challenges presented by the increasing volume of data in graduate and postgraduate programs are effectively addressed by the proposed method. It produces exceptional performance metrics, including a 15-25% decrease in processing time through optimization, 98.7% accuracy, and 0.03% log loss. This work additionally demonstrates the effectiveness of DBTM-SHO in administering extensive datasets and makes an important improvement to educational data mining. It provides a robust foundation for organizations facing the challenges of evaluating student achievement in the era of vast data. [ABSTRACT FROM AUTHOR]

Camoni, the largest digital health community in Israel, involves thousands of patients in the decision-making process concerning their illness and treatment. This approach reflects the recent global shift towards digital tools that combine professional information with social networking capabilities to enable problem-solving, emotional support, and knowledge sharing. Digital health communities serve as an invaluable resource for individuals seeking to learn more about their health, connect with others with shared experiences, and receive encouragement. Our research investigates user trends in digital health communities using the Camoni platform as a case study. To this end, we compile a comprehensive database of 12 years of site activity and conduct a large-scale analysis to identify and assess significant trends in user behavior. We observe several significant trends concerning different genders engagement and note a narrowing of gaps between men and women users' participation and publication volume. Furthermore, we find that younger users have become increasingly active on the platform over time. We also uncover unique gender-specific behavior patterns that we attempt to characterize and explain. Our findings suggest that the rise of digital health communities has accelerated in recent years, reflecting the public's growing preference to take a more active role in their medical care. [ABSTRACT FROM AUTHOR]

Background: The determining effect of facial hard tissues on soft tissue morphology in orthodontic patients has yet to be explained. The aim of this study was to clarify the hard-soft tissue relationships of the lower 1/3 of the face in skeletal Class II-hyperdivergent patients compared with those in Class I-normodivergent patients using network analysis. Methods: Fifty-two adult patients (42 females, 10 males; age, 26.58 ± 5.80 years) were divided into two groups: Group 1, 25 subjects, skeletal Class I normodivergent pattern with straight profile; Group 2, 27 subjects, skeletal Class II hyperdivergent pattern with convex profile. Pretreatment cone-beam computed tomography and three-dimensional facial scans were taken and superimposed, on which landmarks were identified manually, and their coordinate values were used for network analysis. Results: (1) In sagittal direction, Group 2 correlations were generally weaker than Group 1. In both the vertical and sagittal directions of Group 1, the most influential hard tissue landmarks to soft tissues were located between the level of cemento-enamel junction of upper teeth and root apex of lower teeth. In Group 2, the hard tissue landmarks with the greatest influence in vertical direction were distributed more forward and downward than in Group 1. (2) In Group 1, all the correlations for vertical-hard tissue to sagittal-soft tissue position and sagittal-hard tissue to vertical-soft tissue position were positive. However, Group 2 correlations between vertical-hard tissue and sagittal-soft tissue positions were mostly negative. Between sagittal-hard tissue and vertical-soft tissue positions, Group 2 correlations were negative for mandible, and were positive for maxilla and teeth. Conclusion: Compared with Class I normodivergent patients with straight profile, Class II hyperdivergent patients with convex profile had more variations in soft tissue morphology in sagittal direction. In vertical direction, the most relevant hard tissue landmarks on which soft tissue predictions should be based were distributed more forward and downward in Class II hyperdivergent patients with convex profile. Class II hyperdivergent pattern with convex profile was an imbalanced phenotype concerning sagittal and vertical positions of maxillofacial hard and soft tissues. [ABSTRACT FROM AUTHOR]

Predictive modelling in the education domain can be utilised to significantly improve teaching and learning experiences. Massive Open Online Courses (MOOCs) generate a large volume of data that can be exploited to predict and evaluate student performance based on various factors. This paper has two broad aims. Firstly, to develop and tune several Machine Learning (ML) models to perform classification tasks on the dataset to predict student performance, including Linear Regression, Logistic Regression, Random Forests, K-Nearest Neighbours, and more. Secondly, to evaluate the efficacy of these ML models and identify those which are best suited to this task. The categories of data utilised in achieving these aims include (i) demographic information, (ii) academic background, and (iii) interaction with MOOC course materials. The research procedure comprises five phases: data exploration to analyse the dataset, feature engineering which involves discerning the most important features and converting them into a format decipherable by the ML models, model building, model evaluation by measurement of accuracy, and subsequent comparative evaluation between the different models. The results achieved in this study are expected to have implications on how MOOC platforms utilise data to improve user experience. As indicated by the findings of this study, the data collected by these platforms may be used to predict performance with accuracy of over 77%; this extracted information can be exploited to enhance educational theory or practices in the context of MOOCs, for instance by implementing varying teaching methodologies or providing different types of resources based on predicted performance. [ABSTRACT FROM AUTHOR]

In data mining and machine learning communities, Neural Network (NN) is a popular classification method. On extremely unbalanced and complicated datasets, NN may achieve excellent classification accuracy. However, one disadvantage of NN is its inability to explain its reasoning process, which restricts its use in numerous sectors that need clear conclusions as well as high accuracy. To address this issue, rule-extraction mechanisms exist that extract intelligible classification-rules from NN and turn them into a white box. Attribute or network pruning, dealing with diverse attribute types, rule pruning, and dealing with class overlapping difficulties are all significant components or portions of many existing rule extraction methods, and present strategies to deal with these aspects are insufficiently successful. As a result, this study offers a rule extraction approach named "Comprehensible and Transparent Rule Extraction Using Neural Network"-CTRENN to address the aforementioned shortcomings and transform NN into a white box with high accuracy and better explain-ability. The suggested CTRENN is an expansion of the state-of-art Rule Extraction from Neural Network Using Classified and Misclassified Data technique (RxNCM). The CTRENN augments the RxNCM with a floating sequential search for feature and rule selection to improve feature and rule selection. CTRENN also distinguishes between continuous and discrete properties to improve the readability of the produced rules. Unlike RxNCM, the CTRENN employs a probabilistic technique to deal with the overlapping of attribute data ranges in various classes. Experiments are carried out using six real life datasets obtained from the UCI repository in order to illustrate the efficacy of the proposed CTRENN algorithm in comparison to the current methods. [ABSTRACT FROM AUTHOR]

In real-life complex traffic environments, vehicles are often occluded by extraneous background objects and other vehicles, leading to severe degradation of object detector performance. To address this issue, we propose a method named YOLO-OVD (YOLO for occluded vehicle detection) and a dataset for effectively handling vehicle occlusion in various scenarios. To highlight the model attention in unobstructed region of vehicles, we design a novel grouped orthogonal attention (GOA) module to achieve maximum information extraction between channels. We utilize grouping and channel shuffling to address the initialization and computational issues of original orthogonal filters, followed by spatial attention for enhancing spatial features in vehicle-visible regions. We introduce a CIoU-based repulsion term into the loss function to augment the network's localization accuracy in scenarios involving densely packed vehicles. Moreover, we explore the effect of the knowledge-based Laplacian Pyramid on the OVD performance, which contributes to fast convergence in training and ensures more detailed and comprehensive feature retention. We conduct extensive experiments on the established occluded vehicle detection dataset, which demonstrates that the proposed YOLO-OVD model significantly outperforms 14 representative object detectors. Notably, it achieves improvements of 4.7% in Precision, 3.6% in AP@0.5, and 1.9% in AP@0.5:0.95 compared to the YOLOv5 baseline. [ABSTRACT FROM AUTHOR]

Clustering texts together is an essential task in data mining and information retrieval, whose aim is to group unlabeled texts into meaningful clusters that facilitate extracting and understanding useful information from large volumes of textual data. However, clustering short texts (STC) is complex because they typically contain sparse, ambiguous, noisy, and lacking information. One of the challenges for STC is finding a proper representation for short text documents to generate cohesive clusters. However, typically, STC considers only a single-view representation to do clustering. The single-view representation is inefficient for representing text due to its inability to represent different aspects of the target text. In this paper, we propose the most suitable multi-view representation (MVR) (by finding the best combination of different single-view representations) to enhance STC. Our work will explore different types of MVR based on different sets of single-view representation combinations. The combination of the single-view representations is done by a fixed length concatenation via Principal Component analysis (PCA) technique. Three standard datasets (Twitter, Google News, and StackOverflow) are used to evaluate the performances of various sets of MVRs on STC. Based on experimental results, the best combination of single-view representation as an effective for STC was the 5-views MVR (a combination of BERT, GPT, TF-IDF, FastText, and GloVe). Based on that, we can conclude that MVR improves the performance of STC; however, the design for MVR requires selective single-view representations. [ABSTRACT FROM AUTHOR]

The scientific literature on Al-Si alloy materials is a rich source of factual information about various classes of entities (e.g., alloys, properties, and compositions) as well as various relationships between these entities. Automatically extracting this information using natural language processing (NLP) techniques to generate a knowledge graph for Al-Si alloy materials is a challenging task. In this paper, we propose ASaRE-Net (Al-Si alloy Relationship Extraction Network) to extract entities and relationships from Chinese scientific literature on Al-Si alloy materials, thus forming a triple (Entity 1, Relation, Entity 2) as the basic unit for constructing a knowledge graph. Due to the complexity of overlapping triples and many triples in a text sentence, this paper designs a triple-aware module based on knowledge representation learning, which is used to mine the correlation between entities and relations, and uses a simple and efficient labeling strategy to decode, so as to alleviate the problem of cascading errors. This paper defines 11 types of entity and 13 types of relationship for Al-Si alloy materials, and constructs ASaIED (Al-Si alloy Information Extraction Dataset). Experimental results show that compared to the strongest baseline, the proposed model improves the F1 score by 2.34% on the ASaIED dataset, and demonstrates better performance in scenarios involving SEO overlapping types and many triples. The proposed model can continuously provide credible data for constructing knowledge graphs in the field of Al-Si alloy materials. [ABSTRACT FROM AUTHOR]

Extracting information from scanned business documents, while a necessary commercial task, continues to be mostly done manually, requiring significant human effort. Current solutions for automated document information extraction still have limited capabilities in regards to user-required customizability and extraction of dataset-specific information, leaving the area as a very active field of research. In this paper, we propose modifications and improvements to our previously developed custom pipeline for extracting and tabulating key-value pairs from commercial invoice documents. Our design changes and additions adapt the pipeline to a wider variety of document types and use cases, primarily through the implementation of dataset-specific configuration files that promote customizability along with new technical modules that address both general and dataset-specific complexities. We compare our pipeline’s performance against current machine learning and commercial solutions on a real-world dataset, and demonstrate that it is able to extract a wider variety of fields while maintaining competitive or greater accuracies compared to the alternate solutions. [ABSTRACT FROM AUTHOR]

ESG (Environmental, Social and Governance) management practice is an important part of promoting sustainable operation and development of manufacturing enterprises. Currently, traditional evaluation methods have limitations such as low efficiency and lack of objectivity. To improve the efficiency and accuracy of ESG evaluation and promote the optimization of ESG performance in manufacturing enterprises, this article combined data mining and analytic hierarchy process (AHP) to conduct effective research on ESG management practice evaluation in manufacturing enterprises. This article adopted the best priority search strategy to collect and process enterprise ESG data. By using AHP to construct hierarchical and segmented objectives for target problems, a performance evaluation index system for management practices was built based on the evaluation objectives and hierarchical priority order. Finally, based on the performance evaluation of ESG management practices, the K-nearest Neighbor algorithm was applied to analyze historical data of key indicators. According to the weights, various key indicators were re-integrated, achieving practical evaluation and decision support for enterprise ESG management. To verify the effectiveness of data mining and AHP, this article took Z enterprise as the research object and conducted empirical analysis on it. The results showed that in terms of evaluation accuracy, the method proposed in this article achieved the highest evaluation accuracy of 92.51%, 91.16%, and 91.75% in environmental, social, and governance dimension data use case evaluation, respectively. The conclusion indicated that data mining and AHP could improve the accuracy of ESG management practice evaluation in enterprises, provide reliable decision support for enterprise development, and help promote sustainable development of enterprises. [ABSTRACT FROM AUTHOR]

PURPOSE: Patient-centered outcomes (PCOs) are pivotal in cancer treatment, as they directly reflect patients' quality of life. Although multiple studies suggest that factors affecting breast cancer–related morbidity and survival are influenced by treatment side effects and adherence to long-term treatment, such data are generally only available on a smaller scale or from a single center. The primary challenge with collecting these data is that the outcomes are captured as free text in clinical narratives written by clinicians. MATERIALS AND METHODS: Given the complexity of PCO documentation in these narratives, computerized methods are necessary to unlock the wealth of information buried in unstructured text notes that often document PCOs. Inspired by the success of large language models (LLMs), we examined the adaptability of three LLMs, GPT-2, BioGPT, and PMC-LLaMA, on PCO tasks across three institutions, Mayo Clinic, Emory University Hospital, and Stanford University. We developed an open-source framework for fine-tuning LLM that can directly extract the five different categories of PCO from the clinic notes. RESULTS: We found that these LLMs without fine-tuning (zero-shot) struggle with challenging PCO extraction tasks, displaying almost random performance, even with some task-specific examples (few-shot learning). The performance of our fine-tuned, task-specific models is notably superior compared with their non–fine-tuned LLM models. Moreover, the fine-tuned GPT-2 model has demonstrated a significantly better performance than the other two larger LLMs. CONCLUSION: Our discovery indicates that although LLMs serve as effective general-purpose models for tasks across various domains, they require fine-tuning when applied to the clinician domain. Our proposed approach has the potential to lead more efficient, adaptable models for PCO information extraction, reducing reliance on extensive computational resources while still delivering superior performance for specific tasks. New patient-centered outcome extraction pipeline for patients with breast cancer—extensive comparative analysis between LLMs, two external validations, and open-source model sharing. [ABSTRACT FROM AUTHOR]

This paper introduces a novel approach to Indian Sign Language Recognition (ISLR) by integrating Keras, Visual Transformers (ViT), and sophisticated data augmentation techniques. Our methodology emphasizes the development of a Vision Transformer model trained on a comprehensive dataset, enhanced with both image data and keypoint information, leveraging the capabilities of the Mediapipe library. To improve the model's ability to generalize, we applied advanced augmentation strategies, including ImageDataGenerator, among others. Through extensive experimentation, involving rigorous hyperparameter optimization across numerous epochs, we sought to determine the most effective model configuration. The results of our validation process revealed a promising evaluation loss of 0.2941 and an impressive accuracy rate of 97.52%. This integration of data augmentation techniques with ViT transformers establishes a groundbreaking framework in the field of ISLR, underscoring its significant potential for practical implementation and marking a substantial progression in the technology for sign language recognition. [ABSTRACT FROM AUTHOR]

In recent years, with the development of 3D digitization of cultural relics, most cultural sites contain a large number of fine 3D data of cultural relics, especially complex geometric objects such as painted cultural relics. At present, how to automatically extract surface damage information from the fine 3D color model of painted cultural relics and avoid the loss of accuracy caused by reducing the dimension using conventional methods is an urgentproblem. In view of the above issues, this paper proposes an automatic and high-precision extraction method for cultural relics surface shedding diseases based on 3D fine data. First, this paper designs a 2D and 3D integrated data conversion model based on OpenSceneGraph, a 3D engine, which performs mutual conversion between 3D color model textures and 2D images. Second, this paper proposes a simple linear iterative clustering segmentation algorithm with an adaptive k value, which solves the problem of setting the superpixel k value and improves the accuracy of image segmentation. Finally, through the 2D and 3D integrated models, the disease is statistically analyzed and labeled on the 3D model. Experiments show that for painted plastic objects with complex surfaces, the disease extraction method based on the 3D fine model proposed in this paper has improved geometric accuracy compared with the current popular orthophoto extraction method, and the disease investigation is more comprehensive. Compared with the current 3D manual extraction method in commercial software, this method greatly improves the efficiency of disease extraction while ensuring extraction accuracy. The research method of this paper activates many existing 3D fine data of cultural protection units and converts conventional 2D data mining and analysis into 3D, which is more in line with the scientific utilization of data in terms of accuracy and efficiency and has certain scientific research value, leading value and practical significance. [ABSTRACT FROM AUTHOR]

Introduction: The study proposes a method for evaluating stress distribution in high-altitude Tibetan Plateau railway tunnels using high-precision radar satellite time-series interferometric synthetic aperture radar technology. Methods: To effectively monitor and prevent geological hazards during the construction process, this method i employed, as it serves as a component of advanced geological prediction and surrounding rock deformation monitoring technology for high-altitude tunnels, particularly in the Dongelu Tunnel of the China-Tibet Railway. The study utilizes time-series interferometric synthetic aperture radar to obtain deformation information for Dongelu Tunnel area between 2022 and 2023 from Sentinel-1A orbit images. This quantitatively investigates the upper mountain body and line-of-sight direction along the tunnel. The deformation characteristics are correlated with high-frequency and high-precision automated vertical displacement monitoring results, determining the spatiotemporal distribution of tunnel deformation. In this study, a model that determines the vertical stress state of the Dongelu Tunnel under loading near the entrance and evaluates its health status was established. Results: The results show that the surface deformation of the mountain above the tunnel axis develops slowly and is relatively small, with a maximum vertical deformation rate of 1-3 mm/year. The average stress on the tunnel arch is 5.54 MPa, with a fluctuation range of 0.01 MPa. Temporal Q9 changes in various parts of the tunnel are periodic, with maximum fluctuations observed in December 2022. The study reveals inconsistent surface settlement of the tunnel arch and mountain above it, causing minor vertical stress changes. As the tunnel construction progresses, vertical stress variation shows periodicity because of an initial imbalance in internal stress within the mountain. Stress fluctuations near the tunnel entrance occur during the initial excavation phase, gradually diminishing as the project progresses and internal stress stabilizes. Discussion: The proposed tunnel monitoring and stability assessment method can reduce its impact on engineering construction and provide guidance for advanced geological prediction. [ABSTRACT FROM AUTHOR]

Data has become much widely available in recent years. Since the past years, Learning classifiers from unbalanced data is a crucial issue that comes up frequently in classification difficulties. In such cases, the majority of the instances belong to one class while many fewer belong to the other class, which is typically the more significant class. As important data is extracted from data during learning, if the ratio between classes are changed.Class Imbalances causes the classifier's performance to decrease. The imbalanced data issues is well-known in numerous application areas and has recently become a open research challenge in data mining and learning algorithms. In such situation, nearly all the instances belong to majority class, but very few belongs to the minority class, which is often the most crucial class for prediction or detection. Since typical classifications demand a high accuracy over a complete range of examples in this situation, machine learning techniques frequently overwhelm the majority class and neglect the minority class. This survey first evaluated academic initiatives specifically aimed at the issues of class imbalance. Then, we analyzed numerous solutions at four levels during the learning stages. The purpose of survey is to present an overview of class imbalance problem that includes their issues, solution and their disadvantages. The survey concluded with suggestions for future investigation, research problems, and developments in the field. The survey involved the adaptive processes as well. [ABSTRACT FROM AUTHOR]

Objective: Using the Food and Drug Administration Adverse Event Reporting System (FAERS) database, four signal detection methods were applied to mine adverse drug events (ADEs) related to use of dual orexin receptor antagonists (DORAs) to provide reference for safe clinical use. Research design and Methods: Data collected from Q3rd 2014 to Q4th 2023 were obtained from the FAERS database. According to the preferred terminology (PT) and systematic organ classification (SOC) of MedDRA v.26.0, the reporting odds ratio (ROR), proportional reporting ratio (PRR), multi-item gamma Poisson shrinker (MGPS), and Bayesian confidence propagation neural network (BCPNN) were used to detect ADE signals. Results: A total of 11,857 DORAs-related adverse reactions were detected, reported with suvorexant, lemborexant, and daridorexant as the main suspected drugs was 8717584, and 2556, respectively. A higher proportion of females than males were reported (57.27% vs. 33.04%). The top 20 positive PT signals from three DORAs showed that "sleep paralysis" ranked first. "Brain fog" was stronger following daridorexant but was not detected for the other two drugs, and "sleep sex" and "dyssomnia" were stronger in suvorexant but not in the other two drugs. Additionally, some PTs occurred that were not included in drug instructions, such as "hangover" and "hypnagogic hallucination." Conclusion: In this study, four algorithms (ROR, PRR, BCPNN, and MGPS) were used to mine the safety signals of DORAs. We identified some potential ADE signals that can promote the rational use of DORAs and improve their safety. [ABSTRACT FROM AUTHOR]

With the increased availability of chemical data in public databases, innovative techniques and algorithms have emerged for the analysis, exploration, visualization, and extraction of information from these data. One such technique is chemical grouping, where chemicals with common characteristics are categorized into distinct groups based on physicochemical properties, use, biological activity, or a combination. However, existing tools for chemical grouping often require specialized programming skills or the use of commercial software packages. To address these challenges, we developed a user-friendly chemical grouping workflow implemented in KNIME, a free, open-source, low/no-code, data analytics platform. The workflow serves as an all-encompassing tool, expertly incorporating a range of processes such as molecular descriptor calculation, feature selection, dimensionality reduction, hyperparameter search, and supervised and unsupervised machine learning methods, enabling effective chemical grouping and visualization of results. Furthermore, we implemented tools for interpretation, identifying key molecular descriptors for the chemical groups, and using natural language summaries to clarify the rationale behind these groupings. The workflow was designed to run seamlessly in both the KNIME local desktop version and KNIME Server WebPortal as a web application. It incorporates interactive interfaces and guides to assist users in a step-by-step manner. We demonstrate the utility of this workflow through a case study using an eye irritation and corrosion dataset. Scientific contributions This work presents a novel, comprehensive chemical grouping workflow in KNIME, enhancing accessibility by integrating a user-friendly graphical interface that eliminates the need for extensive programming skills. This workflow uniquely combines several features such as automated molecular descriptor calculation, feature selection, dimensionality reduction, and machine learning algorithms (both supervised and unsupervised), with hyperparameter optimization to refine chemical grouping accuracy. Moreover, we have introduced an innovative interpretative step and natural language summaries to elucidate the underlying reasons for chemical groupings, significantly advancing the usability of the tool and interpretability of the results. [ABSTRACT FROM AUTHOR]

Background: Common atypical antipsychotics include risperidone, paliperidone, olanzapine, lurasidone, quetiapine, clozapine, aripiprazole, ziprasidone, asenapine, brexpiprazole, and cariprazine. Previous studies on ocular adverse reactions of antipsychotics were mainly focused on typical antipsychotics. Systematic research on atypical antipsychotics remains limited. Objective: This study aimed to evaluate the potential risks of different atypical antipsychotics causing ocular side effects by mining the Food and Drug Administration Adverse Event Reporting System (FAERS) database. Methods: Extract reports from the FAERS from the first quarter of 2016 to the fourth quarter of 2022 were obtained. Data mining of eye disorders associated with atypical antipsychotics was carried out using The Reporting Odds Ratio (ROR) method and The Medicines and Healthcare Products Regulatory Agency (MHRA) method to determine positive signals. Results: FAERS reports for 9913783 cases were included in these 28 quarters. 64 defined ocular adverse events were classified into 10 categories according to High-Level Group Terms (HLGT). Conclusions: There were differences in the types and severity of ocular-related adverse events associated with atypical antipsychotics. Ocular neuromuscularrelated adverse events were found among all 11 atypical antipsychotics. Olanzapine had the highest signal intensity in oculogyric crisis. Aripiprazole had the highest signal strength in blepharospasm. Cariprazine was associated with cataract-related ocular adverse reactions. In terms of the types of adverse events, our study found that aripiprazole was associated with 28 types of ocular adverse events, followed by quetiapine. Clozapine was only associated with two types of ocular adverse events. [ABSTRACT FROM AUTHOR]

The advancement in the internet and mobile technologies has substantially altered information diffusion in modern society, creating a diverse environment for generating and sharing various forms of information. Specifically, the emergence of new information sources, such as influencers and online communities, has significantly influenced the formation of consumer opinion. We highlight the changes that have occurred in the diffusion of fashion trend information. To do this, we conducted data mining, which involved three main steps: data preprocessing, specifically converting image data (including images from the 2022 F/W season runway collection, fashion influencer outfits, and best items from online fashion retailers) into textual data; data mining analysis (quantitative analysis); and data post-processing. As a result, we found that even items with low or no appearance on the runway held significance in the best item data or fashion influencer outfits. Specifically, the best items on online fashion retailers, reflecting popular fashion trends, had greater similarity to fashion influencer outfits. However, similarities in silhouette attributes were found among runway collections, fashion influencer outfits, and best items data. This study holds great significance because it focuses on fashion items genuinely consumed by the mainstream consumers rather than only focusing on the four major runway collections. Furthermore, these findings offer valuable insights for merchandising and trend forecasting, emphasizing the importance of selectively utilizing fashion trend information in the planning of fashion products. [ABSTRACT FROM AUTHOR]

Most real-time semantic segmentation networks use shallow architectures to achieve fast inference speeds. This approach, however, limits a network's receptive field. Concurrently, feature information extraction is restricted to a single scale, which reduces the network's ability to generalize and maintain robustness. Furthermore, loss of image spatial details negatively impacts segmentation accuracy. To address these limitations, this paper proposes a Multiscale Context Pyramid Pooling and Spatial Detail Enhancement Network (BMSeNet). First, to address the limitation of singular semantic feature scales, a Multiscale Context Pyramid Pooling Module (MSCPPM) is introduced. By leveraging various pooling operations, this module efficiently enlarges the receptive field and better aggregates multiscale contextual information. Moreover, a Spatial Detail Enhancement Module (SDEM) is designed, to effectively compensate for lost spatial detail information and significantly enhance the perception of spatial details. Finally, a Bilateral Attention Fusion Module (BAFM) is proposed. This module leverages pixel positional correlations to guide the network in assigning appropriate weights to the features extracted from the two branches, effectively merging the feature information of both branches. Extensive experiments were conducted on the Cityscapes and CamVid datasets. Experimental results show that the proposed BMSeNet achieves a good balance between inference speed and segmentation accuracy, outperforming some state-of-the-art real-time semantic segmentation methods. [ABSTRACT FROM AUTHOR]

Background: Kidney transplantation is followed by immunosuppressive therapy involving calcineurin inhibitors (CNIs) such as cyclosporin A. However, long-term high CNIs doses can lead to vitamin D deficiency, and genetic variations influencing vitamin D levels can indirectly impact the necessary CNIs dosage. This study investigates the impact of genetic variations of vitamin D binding protein (DBP) rs2282679 and CYP2R1 hydroxylase rs10741657 polymorphisms on the cyclosporin A dosage in kidney transplant recipients. Additional polymorphisims of genes that are predicted to influence the pharmacogenetic profile were included. Methods: Gene polymorphisms in 177 kidney transplant recipients were analyzed using data mining techniques, including the Random Forest algorithm and Classification and Regression Trees (C&RT). The relationship between the concentration/dose (C/D) ratio of cyclosporin A and genetic profiles was assessed to determine the predictive value of DBP rs2282679 and CYP2R1 rs10741657 polymorphisms. Results: Polymorphic variants of the DBP (rs2282679) demonstrated a strong predictive value for the cyclosporin A C/D ratio in post-kidney transplantation patients. By contrast, the CYP2R1 polymorphism (rs10741657) did not show predictive significance. Additionally, the immune response genes rs231775 CTLA4 and rs1800896 IL10 were identified as predictors of cyclosporin A response, though these did not result in statistically significant differences. Conclusions:DBP rs2282679 polymorphisms can significantly predict the cyclosporin A C/D ratio, potentially enhancing the accuracy of CNI dosing. This can help identify patient groups at risk of vitamin D deficiency, ultimately improving the management of kidney transplant recipients. Understanding these genetic influences allows for more personalized and effective treatment strategies, contributing to better long-term outcomes for patients. [ABSTRACT FROM AUTHOR]

Hyperspectral images (HSI) contain abundant spectral information. Efficient extraction and utilization of this information for image classification remain prominent research topics. Previously, hyperspectral classification techniques primarily relied on statistical attributes and mathematical models of spectral data. Deep learning classification techniques have recently been extensively utilized for hyperspectral data classification, yielding promising outcomes. This study proposes a deep learning approach that uses polarization feature maps for classification. Initially, the polar co-ordinate transformation method was employed to convert the spectral information of all pixels in the image into spectral feature maps. Subsequently, the proposed Deep Context Feature Fusion Network (DCFF-NET) was utilized to classify these feature maps. The model was validated using three open-source hyperspectral datasets: Indian Pines, Pavia University, and Salinas. The experimental results indicated that DCFF-NET achieved excellent classification performance. Experimental results on three public HSI datasets demonstrated that the proposed method accurately recognized different objects with an overall accuracy (OA) of 86.68%, 94.73%, and 95.14% based on the pixel method, and 98.15%, 99.86%, and 99.98% based on the pixel-patch method. [ABSTRACT FROM AUTHOR]

Currently, target detection on unmanned aerial vehicle (UAV) images is a research hotspot. Due to the significant scale variability of targets and the interference of complex backgrounds, current target detection models face challenges when applied to UAV images. To address these issues, we designed an effective and lightweight full-scale target detection network, FSTD-Net. The design of FSTD-Net is based on three principal aspects. Firstly, to optimize the extracted target features at different scales while minimizing background noise and sparse feature representations, a multi-scale contextual information extraction module (MSCIEM) is developed. The multi-scale information extraction module (MSIEM) in MSCIEM can better capture multi-scale features, and the contextual information extraction module (CIEM) in MSCIEM is designed to capture long-range contextual information. Secondly, to better adapt to various target shapes at different scales in UAV images, we propose the feature extraction module fitting different shapes (FEMFDS), based on deformable convolutions. Finally, considering low-level features contain rich details, a low-level feature enhancement branch (LLFEB) is designed. The experiments demonstrate that, compared to the second-best model, the proposed FSTD-Net achieves improvements of 3.8%, 2.4%, and 2.0% in AP50, AP, and AP75 on the VisDrone2019, respectively. Additionally, FSTD-Net achieves enhancements of 3.4%, 1.7%, and 1% on the UAVDT dataset. Our proposed FSTD-Net has better detection performance compared to state-of-the-art detection models. The experimental results indicate the effectiveness of the FSTD-Net for target detection in UAV images. [ABSTRACT FROM AUTHOR]

The vision transformer (ViT) has demonstrated performance comparable to that of convolutional neural networks (CNN) in the hyperspectral image classification domain. This is achieved by transforming images into sequence data and mining global spectral-spatial information to establish remote dependencies. Nevertheless, both the ViT and CNNs have their own limitations. For instance, a CNN is constrained by the extent of its receptive field, which prevents it from fully exploiting global spatial-spectral features. Conversely, the ViT is prone to excessive distraction during the feature extraction process. To be able to overcome the problem of insufficient feature information extraction caused using by a single paradigm, this paper proposes an MLP-mixer and a graph convolutional enhanced transformer (MGCET), whose network consists of a spatial-spectral extraction block (SSEB), an MLP-mixer, and a graph convolutional enhanced transformer (GCET). First, spatial-spectral features are extracted using SSEB, and then local spatial-spectral features are fused with global spatial-spectral features by the MLP-mixer. Finally, graph convolution is embedded in multi-head self-attention (MHSA) to mine spatial relationships and similarity between pixels, which further improves the modeling capability of the model. Correlation experiments were conducted on four different HSI datasets. The MGEET algorithm achieved overall accuracies (OAs) of 95.45%, 97.57%, 98.05%, and 98.52% on these datasets. [ABSTRACT FROM AUTHOR]

Deep learning‐based models have become ubiquitous across a wide range of applications, including computer vision, natural language processing, and robotics. Despite their efficacy, one of the significant challenges associated with deep neural network (DNN) models is the potential risk of copyright leakage due to the inherent vulnerability of the entire model architecture and the communication burden of the large models during publishing. So far, it is still challenging for us to safeguard the intellectual property rights of these DNN models while reducing the communication time during model publishing. To this end, this paper introduces a novel approach using knowledge distillation techniques aimed at training a surrogate model to stand in for the original DNN model. To be specific, a knowledge distillation generative adversarial network (KDGAN) model is proposed to train a student model capable of achieving remarkable performance levels while simultaneously safeguarding the copyright integrity of the original large teacher model and improving communication efficiency during model publishing. Herein, comprehensive experiments are conducted to showcase the efficacy of model copyright protection, communication‐efficient model publishing, and the superiority of the proposed KDGAN model over other copyright protection mechanisms. [ABSTRACT FROM AUTHOR]

Multisource domain adaptation (MDA) is committed to mining and extracting data concerning target tasks from several source domains. Many recent studies have focused on extracting domain-invariant features to eliminate domain distribution differences. However, there are three aspects that require further consideration. (1) Efforts should be made to ensure the maximum correlation in the potential subspace between the source and target domains. (2) While aligning the marginal distribution, the conditional distribution must also be considered. (3) Merely aligning the source distribution and target distribution cannot guarantee sufficient differentiation for classification tasks. To address these problems, we propose a novel approach named towards discriminability with distribution discrepancy constrains for multisource domain adaptation (TD-DDC). Specifically, TD-DDC first mines features of maximal relations learned from all domains while constructing domain data distribution mean distance metrics for interdomain distribution adaptation. Simultaneously, we integrate discriminability into domain alignment, which means increasing the distance among labels that are distinct from one another while reducing the distance among labels that are the same. Our proposed method not only reduces the interdomain distributional differences but also takes into account the preservation of interdomain correlation and inter-category discrimination. Numerous experiments have shown that TD-DDC performs much better than its competitors on three visual benchmark test databases. [ABSTRACT FROM AUTHOR]

The development of digital technology, such as data mining and analysis techniques, has enabled e-commerce platforms to use the data generated in their ecosystems and forecast the online demand more accurately. By sharing the forecast information, platforms help their partners reduce the demand uncertainty. To examine the effects of the shared information, this study considers a two-echelon supply chain consisting of one original equipment manufacturer (OEM), one brand store, and one platform, and investigates the relations between the forecast information and firms' channel strategies. Our analysis reveals some interesting implications. First, the platform's forecast information encourages the OEM to develop online business unless the brand store adopts the dual-channel strategy and the platform sets the revenue commission as zero, while it always increases the brand store's willingness to adopt the dual-channel strategy. Moreover, the OEM's online business hinders the platform to share the information if consumers' acceptance for the brand in online channel is low. Interestingly, the brand store's dual-channel strategy hinders the platform to share the information if consumers' acceptance for the brand in online channel is high. Further, we find that when the brand store adopts the dual-channel strategy, the OEM's online business always decreases the benefit of the forecast information for the brand store. However, when the OEM develops online business, the brand store's dual-channel strategy sometimes increases the benefit of the forecast information for the OEM. In addition, we also conduct some numerical experiments to examine the impacts of the platform commission rate on these firms' benefits from the information and find that the higher commission rate can increase the OEM's benefit under certain conditions. Our study aims to provide managerial insights for the OEM, brand store, and platform to capture the value of forecast information. [ABSTRACT FROM AUTHOR]

Climate change is expected to cause important changes in precipitation patterns in Iran until the end of 21st century. This study aims at evaluating projections of climate change over Iran by using five climate model outputs (including ACCESS‐ESM1‐5, BCC‐CSM2‐MR, CanESM5, CMCC‐ESM2 and MRI‐ESM2‐0) of the Coupled Model Intercomparison Project phase 6 (CMIP6), and performing bias‐correction using a novel combination of quantile mapping (QM) and random forest (RF) between the years 2015 and 2100 under three shared socioeconomics pathways (SSP2‐4.5, SSP3‐7.0 and SSP5‐8.5). First, bias‐correction was performed on ERA5‐Land reanalysis data as reference period (1990–2020) using the QM method, then the corrected ERA5‐Land reanalysis data was considered as measured data. Based on the corrected ERA5‐Land reanalysis data (1990–2020) and historical simulations (1990–2014), the future projections (2015–2100) were also bias‐corrected utilizing the QM method. Next, the accuracy of the QM method was validated by comparing the corrected ERA5‐Land reanalysis data with model outputs for overlapping years between 2015 and 2020. This comparison revealed persistent biases; hence, a combination of QM‐RF method was applied to rectify future climate projections until the end of the 21st century. Based on the QM result, CMCC‐ESM2 revealed the highest RMSE in both SSP2‐4.5 and SSP3‐7.0 amounting to 331.74 and 201.84 mm·year−1, respectively. Particularly, the exclusive use of the QM method displayed substantial errors in projecting annual precipitation based on SSP5‐8.5, notably in the case of ACCESS‐ESM1‐5 (RMSE = 431.39 mm·year−1), while the RMSE reduced after using QM‐RF method (197.75 mm·year−1). Obviously, a significant enhancement in results was observed upon implementing the QM‐RF combination method in CMCC‐ESM2 under both SSP2‐4.5 (RMSE = 139.30 mm·year−1) and SSP3‐7.0 (RMSE = 151.43 mm·year−1) showcasing approximately reduction in RMSE values by 192.43 and 50.41 mm·year−1, respectively. Although each bias‐corrected model output was evaluated individually, multi‐model ensemble (MME) was also created to project the annual future precipitation pattern in Iran. By considering that combination of QM‐RF method revealed the lower errors in correcting model outputs, we used the QM‐RF technique to create the MME. Based on SSP2‐4.5, the MME climate projections highlight imminent precipitation reductions (>10%) across large regions of Iran, conversely projecting increases ranging from 10% to over 20% in southern areas under SSP3‐7.0. Moreover, MME projected dramatic declines under SSP5‐8.5, especially impacting central, eastern, and northwest Iran. Notably, the most pronounced possibly decline patterns are projected for arid regions (central plateau) and eastern areas under SSP2‐4.5, SSP3‐7.0 and SSP5‐8.5. [ABSTRACT FROM AUTHOR]

In educational institutions, it is now more important than ever to deliver high-quality academic instruction, and educational data mining is essential for resolving problems that arise from challenging unstructured data in this field. Using machine learning (ML) approaches, the performance of students and traits related to academia, a crucial indicator of higher education, is examined. In the proposed study, the educational dataset is subjected to feature ranking algorithms, including MRMR, ReliefF, Chi-Square, ANOVA, and Kruskal–Wallis, followed by important feature selection using Shapley. The dataset has 16 attributes of integer, categorical type, and after feature ranking approaches, the features with the most important information are chosen and ML techniques are applied to them. It takes two phases to complete the work. The results are obtained after the first phase, in which all features are taken into account for ML training. The second phase of ML training takes into account selective features that are derived using ranking approaches. ML models with only selective attributes are compared to models with all features in order to determine which is more precise. In comparison, the results of the ML models with selective attributes outperformed the models with all attributes. Overall, the ensemblers, i.e., bagged tree and AdaBoost, outperformed other ML techniques such as decision trees, neural networks, naive Bayes, K-nearest neighbor, and support vector machines presented in the proposed study. Bagged trees achieved an accuracy of 81.0 percent, while AdaBoost achieved an accuracy of 74.2 percent. [ABSTRACT FROM AUTHOR]

This paper explores the combination of a data mining-based state evaluation method for electrical equipment in substations, analyzing the effectiveness and accuracy. First, a Gaussian mixture model is applied to fit all raw data of electrical equipment. The Expectation Maximization algorithm summarizes the data distribution characteristics and identifies outliers. The a priori algorithm is then employed for data mining to derive frequent itemsets and association rules between equipment quality and measurement data. For new equipment samples, conditional probabilities of each feature are independently calculated and combined to classify and evaluate equipment quality. The results suggest that equipment reliability in smart substations can be inferred from historical and real-time operational data using improved association rule algorithms and Naive Bayes classifiers. Finally, the proposed method was applied to analyze statistical data from a 110 kV substation of a power supply company. The states prediction accuracy exceeded 95% when compared with actual equipment quality. The effectiveness evaluation metrics demonstrated that this method outperforms single-category algorithms in terms of accuracy and discrimination ability. [ABSTRACT FROM AUTHOR]

Extracting knowledge from cyber threat intelligence is essential for understanding cyber threats and implementing proactive defense measures. However, there is a lack of open datasets in the Chinese cybersecurity field that support both entity and relation extraction tasks. This paper addresses this gap by analyzing vulnerability description texts, which are standardized and knowledge-dense, to create a vulnerability knowledge ontology comprising 13 entities and 15 relations. We annotated 27,311 unique vulnerability description sentences from the China National Vulnerability Database, resulting in a dataset named BVTED for cybersecurity knowledge triple extraction tasks. BVTED contains 97,391 entities and 69,614 relations, with entities expressed in a mix of Chinese and English. To evaluate the dataset's value, we trained five deep learning-based named entity recognition models, two relation extraction models, and two joint entity–relation extraction models on BVTED. Experimental results demonstrate that models trained on this dataset achieve excellent performance in vulnerability knowledge extraction tasks. This work enhances the extraction of cybersecurity knowledge triples from mixed Chinese and English threat intelligence corpora by providing a comprehensive ontology and a new dataset, significantly aiding in the mining, analysis and utilization of the knowledge embedded in cyber threat intelligence. [ABSTRACT FROM AUTHOR]