Information, Volume 13, Issue 10 (October 2022) – 65 articles

Nowadays, people’s lives are filled with a huge amount of picture information, and image retrieval tasks are widely needed. Deep hashing methods are extensively used to manage such demands due to their retrieval rate and memory consumption. The problem with conventional deep hashing image retrieval techniques, however, is that high dimensional semantic content in the image cannot be effectively articulated due to insufficient and unbalanced feature extraction. This paper offers the deep cross-dimensional attention hashing (DCDAH) method considering the flaws in feature extraction, and the important points of this paper are as follows. This paper proposes a cross-dimensional attention (CDA) module embedded in ResNet18; the module can capture the cross-dimension interaction of feature maps to calculate the attention weight effectively because of its special branch. For a feature map acquired by a convolutional neural network (CNN), each branch takes different rotation measurements and residual transformations to process it. To prevent the DCDAH model from becoming too complex, the CDA module is designed to have the characteristics of low computational overhead. This paper introduces a scheme to reduce the dimension of tensors, which can reduce computation and retain abundant representation. For a dimension of a feature map, the Maxpool and Avgpool are performed, respectively, and the two results are connected. The DCDAH method significantly enhances image retrieval performance, according to studies on the CIFAR10 and NUS-WIDE data sets. Full article

Nowadays, web designers are forced to have an even deeper perception of how users approach their products in terms of user experience and usability. Remote Usability Testing (RUT) is the most appropriate tool to assess the usability of web platforms by measuring the level of user attention, satisfaction, and productivity. RUT does not require the physical presence of users and evaluators, but for this very reason makes data collection more difficult. To simplify data collection and analysis and help RUT moderators collect and analyze user’s data in a non-intrusive manner, this research work proposes a low-cost comprehensive framework based on Deep Learning algorithms. The proposed framework, called Miora, employs facial expression recognition, gaze recognition, and analytics algorithms to capture data about other information of interest for in-depth usability analysis, such as interactions with the analyzed software. It uses a comprehensive evaluation methodology to elicit information about usability metrics and presents the results in a series of graphs and statistics so that the moderator can intuitively analyze the different trends related to the KPI used as usability indicators. To demonstrate how the proposed framework could facilitate the collection of large amounts of data and enable moderators to conduct both remote formative and summative tests in a more efficient way than traditional lab-based usability testing, two case studies have been presented: the analysis of an online shop and of a management platform. Obtained results suggest that this framework can be employed in remote usability testing to conduct both formative and summative tests. Full article

Compared with traditional machine learning algorithms, the convolutional neural network (CNN) has an excellent automatic feature learning ability and can complete the nonlinear representation from original data input to output by itself. However, the CNN does not sufficiently mine the tool wear information contained in the multi-sensor data due to disregard of the differences in the contribution of different features when extracting features. In this paper, a tool wear prediction method based on a multi-scale convolutional neural network with attention fusion is proposed, which fuses the tool wear degradation information collected by different types of sensors. In the multi-scale convolution module, convolution kernels with different sizes are used to extract the degradation information of different scales in the wear information, and then the attention fusion module is constructed to fuse the multi-scale feature information. Finally, the mapping between tool wear and multi-sensor data is realized through the feature information obtained by residual connection and full connection layer. By comparing the multi-scale convolutional neural network with different attention mechanisms, the experiments demonstrated the effectiveness and superiority of the proposed method. Full article

With the emergence of a large number of fuzzy spatiotemporal data on the Web, how to represent and operate fuzzy spatiotemporal data has become an important research issue. Meanwhile, the Resource Description Framework (RDF) is a standard data and knowledge description language of the Semantic Web and has been applied in many application areas, such as geographic information systems and meteorological systems. In this paper, a model for representing fuzzy spatiotemporal data is proposed and a set of algebraic operations for the model are investigated. First, a representation method of fuzzy spatiotemporal RDF data and a fuzzy spatiotemporal RDF graph model are proposed. In addition, a formal fuzzy spatiotemporal RDF algebra is proposed and a set of algebraic operations for manipulating fuzzy spatiotemporal RDF data are developed. The algebraic operations include: set operation, selection operation, projection operation, join operation, and construction operation. Finally, the existing SPARQL query language is extended and an example that shows how to apply the proposed algebraic operations to capture the queries expressed by the extended SPARQL query language is given. Full article

Drone delivery has gained increasing importance in the past few years. Recent technology advancements have allowed us to think of systems capable of transporting and delivering goods precisely and efficiently. However, in order to switch from a test environment to a real environment, many open issues need to be addressed. In this paper, we focused on drop-off point localization based on fiducial markers, analyzing different systems and the configuration of different aspects. We tested our system in a real-world environment and drew conclusions which lead us to identify the most reliable fiducial system and family for this use case. Full article

Fact-checking systems have become important tools to verify fake and misguiding news. These systems become more trustworthy when human-readable explanations accompany the veracity labels. However, manual collection of these explanations is expensive and time-consuming. Recent work has used extractive summarization to select a sufficient subset of the most important facts from the ruling comments (RCs) of a professional journalist to obtain fact-checking explanations. However, these explanations lack fluency and sentence coherence. In this work, we present an iterative edit-based algorithm that uses only phrase-level edits to perform unsupervised post-editing of disconnected RCs. To regulate our editing algorithm, we use a scoring function with components including fluency and semantic preservation. In addition, we show the applicability of our approach in a completely unsupervised setting. We experiment with two benchmark datasets, namely LIAR-PLUS and PubHealth. We show that our model generates explanations that are fluent, readable, non-redundant, and cover important information for the fact check. Full article

The classification of biomedical literature is engaged in a number of critical issues that physicians are expected to answer. In many cases, these issues are extremely difficult. This can be conducted for jobs such as diagnosis and treatment, as well as efficient representations of ideas such as medications, procedure codes, and patient visits, as well as in the quick search of a document or disease classification. Pathologies are being sought from clinical notes, among other sources. The goal of this systematic review is to analyze the literature on various problems of classification of medical texts of patients based on criteria such as: the quality of the evaluation metrics used, the different methods of machine learning applied, the different data sets, to highlight the best methods in this type of problem, and to identify the different challenges associated. The study covers the period from 1 January 2016 to 10 July 2022. We used multiple databases and archives of research articles, including Web Of Science, Scopus, MDPI, arXiv, IEEE, and ACM, to find 894 articles dealing with the subject of text classification, which we were able to filter using inclusion and exclusion criteria. Following a thorough review, we selected 33 articles dealing with biological text categorization issues. Following our investigation, we discovered two major issues linked to the methodology and data used for biomedical text classification. First, there is the data-centric challenge, followed by the data quality challenge. Full article

Heterogeneous test processes with respect to test script languages are an integral part of the development process of mechatronic systems that are carried out in supply chains. Up to now, test cases are not exchangeable between test processes because interoperability is not given. The developed approach enables the source-to-source compiling of test cases between test script languages. With this, the interoperability of test cases is achieved, and seamless integration within the supply chain is possible. The developed approach uses transcompilers as a baseline. In doing so, an interoperability model for test cases is presented. Based on the interoperability model, a source-to-source compiling for test cases is shown. The outcome is a prototype that handles test script languages, which are different with respect to type safety and applied programming paradigms. The approach ensures that test cases are still understandable and usable for test reports. The evaluation confirms the translation capabilities as well as the readability of the generated test case for the high-lift scenario from aviation. The interoperability of test cases within the supply chain enables the formalisation of procedural test knowledge to be used in a broad range of future scenarios, such as test automation, digital twins and predictive maintenance. Full article

Accurate sales forecasting can provide a scientific basis for the management decisions of enterprises. We proposed the xDeepFM-LSTM combined forecasting model for the characteristics of sales data of apparel retail enterprises. We first used the Extreme Deep Factorization Machine (xDeepFM) model to explore the correlation between the sales influencing features as much as possible, and then modeled the sales prediction. Next, we used the Long Short-Term Memory (LSTM) model for residual correction to improve the accuracy of the prediction model. We then designed and implemented comparison experiments between the combined xDeepFM-LSTM forecasting model and other forecasting models. The experimental results show that the forecasting performance of xDeepFM-LSTM is significantly better than other forecasting models. Compared with the xDeepFM forecasting model, the combined forecasting model has a higher optimization rate, which provides a scientific basis for apparel companies to make adjustments to adjust their demand plans. Full article

Handwritten signatures have traditionally been used as a common form of recognition and authentication in tasks such as financial transactions and document authentication. However, there are few studies on minority languages such as Uyghur and Kazakh used in Xinjiang, China, and no available public dataset for these scripts, which are widely used in banking and other fields. Therefore, this paper addresses this problem by constructing a dataset containing Uyghur, Kazakh, and Han languages and presents an automatic handwritten signature recognition approach based on Uyghur, Kazakh, Han, and public datasets. In the paper, a handwritten signature recognition method that combines local maximum occurrence features (LOMO) and histogram of orientated gradients (HOG) features was proposed. LOMO features use a sliding window to represent the local features of the signature image. The high-dimensional features formed by the combination of these methods are dimensionally reduced by principal component analysis (PCA). The classification is performed using k-nearest neighbors (k-NN), and it is compared with the random forest method. The proposed method achieved a recognition rate of 98.4% using a diverse signature database compared with existing methods. It shows that the method was effective and can be applied to large datasets of mixed, multilingual signatures. Full article

A holistic occupational and economy-wide framework for salary prediction is developed and tested using statistical machine learning (ML). Predictive models are developed based on occupational features and organizational characteristics. Five different supervised ML algorithms are trained using survey data from the Saudi Arabian labor market to estimate mean annual salary across economic activities and major occupational groups. In predicting the mean salary over economic activities, the Bayesian Gaussian process regression ML showed a marked improvement in $R^2$ over multiple linear regression (from 0.50 to 0.98). Moreover, lower error levels were obtained: root-mean-square error was reduced by 80% and mean absolute error was reduced by almost 90% compared to multiple linear regression. However, the salary prediction over major occupational groups resulted in artificial neural networks performing the best in terms of both $R^2$, with an improvement from 0.62 in multiple linear regression to 0.94 and errors were reduced by approximately 60%. The proposed framework can help estimate annual salary levels across different types of economic activities and organization sizes, as well as different occupations. Full article

In general, automation involves less human intervention, which leads to dependence on preprogrammed machines and processes that operate continually and carry out numerous tasks. This leads to predictable repeating behavior that can be used to advantage. Due to the incorporation of the Internet of Things into such automated processes, these cyber–physical systems are now vulnerable to cyberattacks, the patterns of which can be difficult to identify and understand. Wastewater treatment plants (WTPs) can be challenging to run, but the treatment process is essential since drinking water and water that can be recycled are extremely important to obtain. The increasing susceptibility of WTPs to cyberattacks brought on by exploitation of their weaknesses poses a further challenge. Understanding system weaknesses and potential exploits is necessary for securing such cyber–physical systems. An attack graph utilization and visualization approach for WTPs is presented in this paper. A formal modeling and encoding of the system were carried out using a structural framework (AADL). The system model was then continuously checked by a model-checker called JKind against security requirements to create attack routes, which were then merged into an attack graph using a tool called GraphViz. Full article

Obtaining soil water content and soil workability data using remote sensing technology with passive sensors has some limitations due to cloud cover, cloud shadow, haze and smoke. This study proposes a method for computing soil water content and soil workability over large areas, faster and in near real-time based on Sentinel-1A (SAR) data. Sample data collected from sugarcane plantations in the Kediri and Sidoarjo districts in East Java, Indonesia, were used to develop a mathematical model of the proposed method using multi-polynomial regression. The performance indicators of the model (RMSE, MAPE and accuracy) were calculated with the results of RMSE = 0.213 and 0.250, MAPE = 16.39% and 18.79%, and accuracy = 83.6% and 81.2% for the training and testing models, respectively. The distribution of soil water content and soil workability can be computed and visualized using a spatial map. The future contribution of this work is to develop a decision support system for the selection of appropriate machinery for sugarcane field operations based on the principles of precision agriculture. Full article

In multi-lingual, multi-speaker environments (e.g., international conference scenarios), speech, language, and background sounds can overlap. In real-world scenarios, source separation techniques are needed to separate target sounds. Downstream tasks, such as ASR, speaker recognition, speech recognition, VAD, etc., can be combined with speech separation tasks to gain a better understanding. Since most of the evaluation methods for monophonic separation are either single or subjective, this paper used the downstream recognition task as an overall evaluation criterion. Thus, the performance could be directly evaluated by the metrics of the downstream task. In this paper, we investigated a two-stage training scheme that combined speech separation and language identification tasks. To analyze and optimize the separation performance of single-channel overlapping speech, the separated speech was fed to a language identification engine to evaluate its accuracy. The speech separation model was a single-channel speech separation network trained with WSJ0-2mix. For the language identification system, we used an Oriental Language Dataset and a dataset synthesized by directly mixing different proportions of speech groups. The combined effect of these two models was evaluated for various overlapping speech scenarios. When the language identification network model was based on single-person single-speech frequency spectrum features, Chinese, Japanese, Korean, Indonesian, and Vietnamese had significantly improved recognition results over the mixed audio spectrum. Full article

Attempts to detect socially desirable responding bias have mainly focused on studies that explore sensitive topics. However, researchers concur that the sensitive character of the survey could be affected by the social context within which the research is conducted. Little research has been reported worldwide investigating the potential effects of social desirability on students’ self-reports, considering the social context within which the survey is conducted. In this paper, we investigate the potential effects of social desirability on students’ self-reports in two social contexts within which the survey was conducted. More specifically, with a sample of 111 Greek students, we explored the effects of social desirability on students’ attitudes towards statistics in two cases: when the questionnaire was administered to participating students after attending (a) lectures and (b) both lectures and laboratory classes. Only in the second case were the items’ attitudes toward statistics associated with a score of socially desirable responding; moreover, social desirability accounted for the relationship between attitudes toward statistics and perceived competence in mathematics. Implications and limitations are also discussed. Full article

Public virtual profiles arose with the evolution of the web and its related technologies. The individual virtual profiles leave a digital footprint that serves as a showcase of the individual. The analysis and management of what is known as digital identity should be an element to be mastered within the digital competencies of future professionals and current university students. This work describes the research carried out over four years through the Digital Identity Workshop, whose public is higher education students. The research has a double objective; first, to study the student’s self-analysis and self-reflection based on his presence on the web; second, to learn strategies for correctly managing his digital identity from the professional and academic point of view. The result has been a success in meeting these objectives after the various editions of the workshop. Pre and post-tests show a significant increase in the students’ digital skills in this field of personal branding. Full article

The responsibility and role of human drivers during automated driving might change dynamically. In such cases, human-machine interface (HMI) transparency becomes crucial to facilitate driving safety, as the states of the automated vehicle have to be communicated correctly and efficiently. However, there is no standardized transparency assessment method to evaluate the understanding of human drivers toward the HMI. In this study, we defined functional transparency (FT) and, based on this definition, proposed a transparency assessment method as a preliminary step toward the objective measurement for HMI understanding. The proposed method was verified in an online survey where HMIs of different vehicle manufacturers were adopted and their transparencies assessed. Even though no significant result was found among HMI designs, FT was found to be significantly higher for participants more experienced with SAE Level 2 automated vehicles, suggesting that more experienced users understand the HMIs better. Further identification tests revealed that more icons in BMW’s and VW’s HMI designs were correctly used to evaluate the state of longitudinal and lateral control. This study provides a novel method for assessing transparency and minimizing confusion during automated driving, which could greatly assist the HMI design process in the future. Full article

The analysis of influential machine parameters can be useful to plan and design a plastic injection molding process. However, current research in parameter analysis is mostly based on computer-aided engineering (CAE) or simulation which have been demonstrated to be inadequate for analyzing complex behavioral changes in the real injection molding process. More advanced approaches using machine learning technology specifically with artificial neural networks (ANNs) brought promising results in terms of prediction accuracy. Nevertheless, the black box and distributed representation of ANN prevent humans from gaining an insight into which process parameters give a significant influence on the final prediction output. Therefore, in this paper, we develop a simpler ANN model by using structural learning with forgetting (SLF) as the algorithm for the training process. Instead of typical backpropagation which generated a fully connected layer of the ANN model, SLF only reveals the important neurons and connections. Hence, the training process of SLF leaves only influential connections and neurons. Since each of the neurons specifically on the input layer represent each of the injection molding parameters, the ANN-SLF model can be further investigated to determine the influential process parameters. By applying SLF to the ANN training process, this experiment has successfully extracted a set of significant injection molding process parameters. Full article

Life-critical embedded systems, including medical devices, are becoming increasingly interconnected and interoperable, providing great efficiency to the healthcare ecosystem. These systems incorporate complex software that plays a significantly integrative and critical role. However, this complexity substantially increases the potential for cybersecurity threats, which directly impact patients’ safety and privacy. With software continuing to play a fundamental role in life-critical embedded systems, maintaining its trustworthiness by incorporating fail-safe modes via a multimodal design is essential. Comprehensive and proactive evaluation and management of cybersecurity risks are essential from the very design to deployment and long-term management. In this paper, we present FIRE, a finely integrated risk evaluation methodology for life-critical embedded systems. Security risks are carefully evaluated in a bottom-up approach from operations-to-system modes by adopting and expanding well-established vulnerability scoring schemes for life-critical systems, considering the impact to patient health and data sensitivity. FIRE combines a static risk evaluation with runtime dynamic risk evaluation to establish comprehensive risk management throughout the lifecycle of the life-critical embedded system. We demonstrate the details and effectiveness of our methodology in systematically evaluating risks and conditions for risk mitigation with a smart connected insulin pump case study. Under normal conditions and eight different malware threats, the experimental results demonstrate effective threat mitigation by mode switching with a 0% false-positive mode switching rate. Full article

Digital transformation of the healthcare industry is being accelerated due to the evolution of digital intelligence information technology such as artificial intelligence (AI), Internet of Things (IoT), and big data. As online-to-offline (O2O)-based consumption life, based on platforms, becomes routinized along with the COVID-19 pandemic, the O2O platforms on medical activities are gaining attention. This study targeted the medical information O2O platform users and aimed to verify the effects of service quality factors on the platform users’ continuous use intention with the mediation of perceived usefulness and perceived convenience. Based on previous studies, four such factors: context-based affordability, immediacy of connection, reliability, and safety were defined as the medical information O2O platform service quality components. This study targeted 369 users of medical information O2O platforms with market dominance in Korea and conducted a questionnaire survey. According to analysis results, context-based affordability and immediacy of connection had a positive (+) effect on perceived usefulness and convenience, and they were confirmed to affect continuous-use intention with the mediation of the perceived usefulness and convenience. Meanwhile, reliability did not affect the perceived usefulness and convenience, whereas safety had a positive (+) effect on perceived usefulness but did not have the same effect (+) on perceived convenience. Consequently, it was ascertained that context-based affordability and immediacy of connection are more important factors to the medical information O2O platform consumers than reliability and safety. Full article

Gestational diabetes mellitus (GDM) is a common condition with repercussions for both the mother and her child. Machine learning (ML) modeling techniques were proposed to predict the risk of several medical outcomes. A systematic evaluation of the predictive capacity of maternal factors resulting in GDM in the UAE is warranted. Data on a total of 3858 women who gave birth and had information on their GDM status in a birth cohort were used to fit the GDM risk prediction model. Information used for the predictive modeling were from self-reported epidemiological data collected at early gestation. Three different ML models, random forest (RF), gradient boosting model (GBM), and extreme gradient boosting (XGBoost), were used to predict GDM. Furthermore, to provide local interpretation of each feature in GDM diagnosis, features were studied using Shapley additive explanations (SHAP). Results obtained using ML models show that XGBoost, which achieved an AUC of 0.77, performed better compared to RF and GBM. Individual feature importance using SHAP value and the XGBoost model show that previous GDM diagnosis, maternal age, body mass index, and gravidity play a vital role in GDM diagnosis. ML models using self-reported epidemiological data are useful and feasible in prediction models for GDM diagnosis amongst pregnant women. Such data should be periodically collected at early pregnancy for health professionals to intervene at earlier stages to prevent adverse outcomes in pregnancy and delivery. The XGBoost algorithm was the optimal model for identifying the features that predict GDM diagnosis. Full article

With the increasing popularity of Twitter as both a social media platform and a data source for companies, decision makers, advertisers, and even researchers alike, data have been so massive that manual labeling is no longer feasible. This research uses a semi-supervised approach to sentiment analysis of both English and Tagalog tweets using a base classifier. In this study involving the Philippines, where social media played a central role in the campaign of both candidates, the tweets during the widely contested race between the son of the Philippines’ former President and Dictator, and the outgoing Vice President of the Philippines were used. Using Natural Language Processing techniques, these tweets were annotated, processed, and trained to classify both English and Tagalog tweets into three polarities: positive, neutral, and negative. Through the Self-Training with Multinomial Naïve Bayes as base classifier with 30% unlabeled data, the results yielded an accuracy of 84.83%, which outweighs other studies using Twitter data from the Philippines. Full article

The existing forward collision warning (FCW) systems that adopt kinematic or perceptual parameters have some drawbacks in the warning performance because of poor adaptability to the users or ineffectiveness of the warnings. To solve the problems of adaptability, several FCW models have been proposed based on algorithms (machine learning, deep learning). However, there is a lack of consideration for the multi-staged warning to avoid an abrupt warning that may startle or distract the driver. In this study, a light gradient boosting machine (LGBM) was adopted to develop a multi-staged FCW. The proposed model was trained and evaluated on a platform based on a driving simulator by twenty drivers. Through Shapley Additive Explanations (SHAPs), the output of the proposed model was explained. Specifically, the front vehicle acceleration, time-to-collision (TTC), and relative speed were found to strongly affect the warning stages from the proposed model. To evaluate the utility and acceptability of the developed model, it was compared with three existing FCW models in terms of subjective and objective indicators. As a result, a trade-off was found between the utility and user acceptance. Additionally, the comparison study also indicated that the developed model outperformed other previous models due to not only the high accuracy but also the suitable trigger timing for each participant. Full article

The article presents an intention to examine the possibilities of processing data on the representation of women in science and research from data collected in Slovakia as part of the Gender Equality Plan. The methodology follows the declared intention and consists of three steps. The first step is the identification of sources of sex-disaggregated data from the field of science and research in the Slovak Republic. Then follows the examination of the state of the art of tracking data in the identified data sources. The analysis of available data and the processing of the results is the next step. The share of women in Slovak science and research is demonstrated by the composition of project teams and by the statistical data of the supplementary statistical survey of research and development potential, which are collected through the national information system for research, development, and innovation, named SK CRIS. The result is a detailed analysis of the position of women in Slovak science and research, classified by research area and academic career stage. Based on the research conducted and the results achieved, we underline the importance of building national information systems in science and research. Data from these systems can significantly contribute to the creation and parameterization of science policy, including the principles of gender equality. Full article

At present, secure multi-party computing is an effective solution for organizations and institutions that want to derive greater value and benefit from the collaborative computing of their data. Most current secure multi-party computing solutions use encryption schemes that are not resistant to quantum attacks, which is a security risk in today’s quickly growing quantum computing, and, when obtaining results, the result querier needs to collect the private keys of multiple data owners to jointly decrypt them, or there needs to be an interaction between the data owner and the querier during the decryption process. Based on the NTRU cryptosystem, which is resistant to quantum computing attacks and has a simple and easy-to-implement structure, and combined with multi-key fully homomorphic encryption (MKFHE) and proxy re-encryption, this paper proposes a secure multi-party computing scheme based on NTRU-type multi-key fully homomorphic proxy re-encryption in the blockchain environment, using the blockchain as trusted storage and a trusted execution environment to provide data security for multi-party computing. The scheme meets the requirements of being verifiable, conspiracy-proof, individually decryptable by the querier, and resistant to quantum attacks. Full article

Driving automation deeply modifies the role of the human operator behind the steering wheel. Trust is required for drivers to engage in such automation, and this trust also seems to be a determinant of drivers’ behaviors during automated drives. On the one hand, first experiences with automation, either positive or not, are essential for drivers to calibrate their level of trust. On the other hand, an automation that provides feedback about its own level of capability to handle a specific driving situation may also help drivers to calibrate their level of trust. The reported experiment was undertaken to examine how the combination of these two effects will impact the driver trust calibration process. Four groups of drivers were randomly created. Each experienced either an early (i.e., directly after the beginning of the drive) or a late (i.e., directly before the end of it) critical situation that was poorly handled by the automation. In addition, they experienced either a consistent continuous feedback (i.e., that always correctly informed them about the situation), or an inconsistent one (i.e., that sometimes indicated dangers when there were none) during an automated drive in a driving simulator. Results showed the early- and poorly-handled critical situation had an enduring negative effect on drivers’ trust development compared to drivers who did not experience it. While being correctly understood, inconsistent feedback did not have an effect on trust during properly managed situations. These results suggest that the performance of the automation has the most severe influence on trust, and the automation’s feedback does not necessarily have the ability to influence drivers’ trust calibration during automated driving. Full article

Business-based decision support systems have been proposed for a few decades in the e-commerce and textile industries. However, these Decision Support Systems (DSS) have not been so productive in terms of business decision delivery. In our proposed model, we introduce a content-based image retrieval model based on a DSS and recommendations system for the textile industry, either offline or online. We used the Fashion MNIST dataset developed by Zalando to train our deep learning model. Our proposed hybrid model can demonstrate how a DSS can be integrated with a system that can separate customers based on their personal characteristics in order to tailor recommendations of products using behavioral analytics, which is trained based on MBTI personality data and Deap EEG data containing numerous pre-trained EEG brain waves. With this hybrid, a DSS can also show product usage analytics. Our proposed model has achieved the maximum accuracy compared to other proposed state-of-the-art models due to its qualitative analysis. In the first section of our analysis, we used a deep learning algorithm to train our CBIR model based on different classifiers such as VGG-net, Inception-Net, and U-net which have achieved an accuracy of 98.2% with a 2% of minimized error rate. The result was validated using different performance metrics such as F-score, F-weight, Precision, and Recall. The second part of our model has been tested on different machine learning algorithms with an accuracy rate of 89.9%. Thus, the entire model has been trained, validated, and tested separately to gain maximum efficiency. Our proposal for a DSS system, which integrates several subsystems with distinct functional sets and several model subsystems, is what makes this study special. Customer preference is one of the major problems facing merchants in the textile industry. Additionally, it can be extremely difficult for retailers to predict customer interests and preferences to create products that fulfill those needs. The three innovations presented in this work are a conceptual model for personality characterization, utilizing an amalgamation of an ECG classification model, a suggestion for a textile image retrieval model using Denoising Auto-Encoder, and a language model based on the MBTI for customer rating. Additionally, we have proposed a section showing how blockchain integration in data pre-processing can enhance its security and AI-based software quality assurance in a multi-model system. Full article

The introduction of game-based learning techniques has significantly swayed learning, motivation, and information processing in both traditional and digital learning environments. This paper studies the footprint that the implementation of ten short-duration digital escape rooms has had on the creation of an environment of positive emotions in the educational field. The digital escape rooms were created by employing the Genial.ly platform and RPG Maker MZ software. A feelings/satisfaction questionnaire has been conducted to study what emotions students have experienced, as well as the students’ opinions about essential elements of digital escape rooms, to study whether positive feelings predominate in the performance of these activities. Results show a high incidence of positive emotions, and a very favorable opinion on the tools employed and the positive feelings on the acquisition of knowledge and skills. Full article

Clustering analysis is a significant technique in various fields, including unsupervised machine learning, data mining, pattern recognition, and image analysis. Many clustering algorithms are currently used, but almost all of them encounter various challenges, such as low accuracy, required number of clusters, slow processing, inability to produce non-spherical shaped clusters, and unstable performance with respect to data characteristics and size. In this research, a novel clustering algorithm called the fast component density clustering in spatial databases (FCDCSD) is proposed by utilizing a density-based clustering technique to address the aforementioned existing challenges. First, from the smallest to the largest point in the spatial field, each point is labeled with a temporary value, and the adjacent values in one component are stored in a set. Then, all sets with shared values are merged and resolved to obtain a single value that is representative of the merged sets. These values represent final cluster values; that is, the temporary equivalents in the dataset are replaced to generate the final clusters. If some noise appears, then a post-process is performed, and values are assigned to the nearest cluster based on a set of rules. Various synthetic datasets were used in the experiments to evaluate the efficiency of the proposed method. Results indicate that FCDCSD is generally superior to affinity propagation, agglomerative hierarchical, k-means, mean-shift, spectral, and density-based spatial clustering of applications with noise, ordering points for identifying clustering structures, and Gaussian mixture clustering methods. Full article