Adequate Data Sets Research Articles

Indonesian Fake News Classification Using Transfer Learning in CNN and LSTM

Fake news spreads quickly and is challenging to stop due to the ease of accessing and sharing information online. Deep learning techniques are a method that can be used to identify fake news quickly and accurately. The types of neural networks commonly utilized in deep learning architectures include Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM), which can perform well when managing the task of classifying fake news, according to several pertinent studies. Regarding handling instances of Indonesian fake news classification, this study compares how well the CNN and LSTM models perform. However, given that Indonesian is a low-resource language with scant documentation, it is challenging to build an adequate data set. At the same time, the CNN and LSTM classification models require significant training data. We proposed a transfer learning method by combining two classification models with a pre-trained IndoBERT language model. 1340 news text data were used, including 643 actual news texts from CNN Indonesia, Liputan6, and Detik and 697 fake news texts from TurnBackHoax. As a result, the performance of the combination of the LSTM classification model with IndoBERT outperformed that of the CNN classification model with IndoBERT, which only produced an accuracy of 92.91%, down by 6%, and was able to produce an accuracy of up to 97.76%, an increase of 4.8% from before. Furthermore, the results show that the LSTM classification model outperforms the CNN classification model in capturing the representation created by IndoBERT. Additionally, these insights may serve as a basis for future research on identifying fake news in Indonesia, helping to improve methods for combatting misinformation in Indonesia.

A novel multi-scale violence and public gathering dataset for crowd behavior classification

Dependable utilization of computer vision applications, such as smart surveillance, requires training deep learning networks on datasets that sufficiently represent the classes of interest. However, the bottleneck in many computer vision applications lies in the limited availability of adequate datasets. One particular application that is of great importance for the safety of cities and crowded areas is smart surveillance. Conventional surveillance methods are reactive and often ineffective in enable real-time action. However, smart surveillance is a key component of smart and proactive security in a smart city. Motivated by a smart city application which aims at the automatic identification of concerning events for alerting law-enforcement and governmental agencies, we craft a large video dataset that focuses on the distinction between small-scale violence, large-scale violence, peaceful gatherings, and natural events. This dataset classifies public events along two axes, the size of the crowd observed and the level of perceived violence in the crowd. We name this newly-built dataset the Multi-Scale Violence and Public Gathering (MSV-PG) dataset. The videos in the dataset go through several pre-processing steps to prepare them to be fed into a deep learning architecture. We conduct several experiments on the MSV-PG datasets using a ResNet3D, a Swin Transformer and an R(2 + 1)D architecture. The results achieved by these models when trained on the MSV-PG dataset, 88.37%, 89.76%, and 89.3%, respectively, indicate that the dataset is well-labeled and is rich enough to train deep learning models for automatic smart surveillance for diverse scenarios.

An end-to-end learning approach for enhancing intrusion detection in Industrial-Internet of Things

The Industrial-Internet of Things (I-IoT) stands out as one of the most dynamically evolving subfields within the expansive realm of the Internet of Things (IoT). Its exponential growth is reshaping industrial landscapes, bringing forth transformative innovations and advancements at an unprecedented pace, as the core of Industry 4.0. Among the formidable challenges faced by the Industrial-Internet of Things, cybersecurity stands out as a critical concern. Deep learning-based Intrusion Detection System (IDS) solutions showcase their steadfast ability to secure resource-limited, investigation-demanding, and complex I-IoT environments. However, their effectiveness hinges not only on the model but also on the dataset on which they are trained. While numerous literature studies delve into this field, existing proposed models often grapple with challenges. They are frequently trained on outdated, non-diverse datasets or lack specific features crucial for I-IoT networks. Recent efforts, thankfully, introduce more adequate datasets like Edge-IIoTset. Researchers leverage this extensive dataset to train models, focusing on detecting the 14 sophisticated attacks. These attacks predominantly target real I-IoT networks. Despite these efforts, none of the existing models proves entirely efficient. A review of literature solutions reveals that many models cannot detect all 15 classes in the dataset. Some are multi-staged or overly complex. In response to these challenges, this paper presents an End-to-End learning , non-complex CNN1D model tailored to the specific problem of detecting 14 sophisticated threats targeting I-IoT environments. Our proposed model demonstrated remarkable efficiency with an accuracy of 99.96%, successfully detecting all 15 classes in the Edge-IIoTset dataset with a minimal loss of 0.0011. Not only that, but our model was validated with k-fold cross-validation, demonstrating its efficiency in preserving the same performance on unseen data and its ability to be generalized for real-world I-IoT environments.

Contributions to a global understanding of socioenvironmental justice related to urban forest: Trends from Brazilian cities in the southeastern Paraná State

Global studies on urban forests (UFs) have extensively covered both environmental and social aspects, including concerns about public health and green solutions provided by trees and forests in cities. However, this trend has not been widely observed in South America, particularly in cities located outside and distant from capital cities and wealthier regions. So, this study examines disparities in socioenvironmental justice of cities in the poorest region of Paraná State, Brazil, comparing to Curitiba (CWB), the capital and one of the greenest cities of Brazil, and proposes a methodology to surpass the scarcity of adequate dataset. A four-dimensional analysis covers the configuration of UF landscapes, the provision of ecosystem services, accessibility to UF, and their interrelation with socioeconomic indicators. We used high-resolution multispectral images from PlanetScope Dove and employed a machine learning algorithm to accurately map UF. We quantified landscape patterns of UFs using a variety of landscape metrics (LM), including mean patch area (AREA_MN), percentage of the landscape class (PLAND), largest patch index (LPI), mean shape index (SHAPE_MN), and the percentage of similar adjacencies (PLADJ). Additionally, we estimated ecosystem services such as total biomass (TB), carbon stock (TCS), carbon dioxide equivalent (CO2eq), urban heat island (UHI), and available public green areas (PGA). We created accessibility maps to PGA at distances of 300 and 800 m. A Mann-Kendall correlation analysis was employed to explore the relationships between these variables and various socioeconomic indicators, including the percentage of people in households without adequate water supply and sewage (TAEE), the percentage of the population living in urban households without garbage collection service (TPSL), the illiteracy rate of the population aged 15 years or older (TANAL), the proportion of people with per capita household income equal to or less than half minimum wage at year of 2010 (TVULNER), the number of schools (SCHOOLS), and the number of permanent residents (MPERM). The Sen's slope test was performed to determine the magnitude of these correlations. Our four-dimensional analysis revealed significant disparities in the distribution and accessibility of ecosystem services within the cities of southeastern Paraná State when compared to CWB. Correlation analyses revealed significant discrepancies between different socioeconomic conditions and the availability of ecosystem services. Financial vulnerability is positively correlated with ecosystem services, while the lack of sanitation negatively impacts their availability. The results highlight the urgent need for equitable distribution of UF ecosystem services.

QuerySum: A Multi-Document Query-Focused Summarization Dataset Augmented with Similar Query Clusters

Query-focused summarization (QFS) aims to summarize the source document(s) with regard to a specific aspect of information given in a query. It plays an important role in presenting users with a concise answer summary from a set of query-relevant documents retrieved by the information retrieval system. Nonetheless, the QFS research has long been hampered by the lack of adequate datasets in terms of both quality and quantity. In this paper, we introduce a large-scale multi-document query-focused summarization dataset, called QuerySum, which contains 27,041 data samples covering diverse topics and its quality is guaranteed through human verification. Unlike some previous QFS datasets constructed directly from the question answering datasets, 74% queries in our dataset are the challenging non-factoid What-, Why-, and How- questions. More importantly, we also provide a set of similar queries together with the corresponding summaries pairs for each query as the retrieved context, presenting a new feature of QuerySum. We aim to encourage research efforts in query intention understanding in the context of QFS. Leveraging QuerySum's depth, we propose a model for query-aware multi-document summarization and set a new QFS benchmark.

The improved stratified transformer for organ segmentation of Arabidopsis.

Segmenting plant organs is a crucial step in extracting plant phenotypes. Despite the advancements in point-based neural networks, the field of plant point cloud segmentation suffers from a lack of adequate datasets. In this study, we addressed this issue by generating Arabidopsis models using L-system and proposing the surface-weighted sampling method. This approach enables automated point sampling and annotation, resulting in fully annotated point clouds. To create the Arabidopsis dataset, we employed Voxel Centroid Sampling and Random Sampling as point cloud downsampling methods, effectively reducing the number of points. To enhance the efficiency of semantic segmentation in plant point clouds, we introduced the Plant Stratified Transformer. This network is an improved version of the Stratified Transformer, incorporating the Fast Downsample Layer. Our improved network underwent training and testing on our dataset, and we compared its performance with PointNet++, PAConv, and the original Stratified Transformer network. For semantic segmentation, our improved network achieved mean Precision, Recall, F1-score and IoU of 84.20, 83.03, 83.61 and 73.11%, respectively. It outperformed PointNet++ and PAConv and performed similarly to the original network. Regarding efficiency, the training time and inference time were 714.3 and 597.9 ms, respectively, which were reduced by 320.9 and 271.8 ms, respectively, compared to the original network. The improved network significantly accelerated the speed of feeding point clouds into the network while maintaining segmentation performance. We demonstrated the potential of virtual plants and deep learning methods in rapidly extracting plant phenotypes, contributing to the advancement of plant phenotype research.

High Accuracy Location Information Extraction From Social Network Texts Using Natural Language Processing

Terrorism has become a worldwide plague with severe consequences for the development of nations. Besides killing innocent people daily and preventing educational activities from taking place, terrorism is also hindering economic growth. Machine Learning (ML) and Natural Language Processing (NLP) can contribute to fighting terrorism by predicting in real-time future terrorist attacks if accurate data is available. This paper is part of a research project that uses text from social networks to extract necessary information to build an adequate dataset for terrorist attack prediction. We collected a set of 3000 social network texts about terrorism in Burkina Faso and used a subset to experiment with existing NLP solutions. The experiment reveals that existing solutions have poor accuracy for location recognition, which our solution resolves. We will extend the solution to extract dates and action information to achieve the project's goal.

Prediction of the superimposed laser shot number for copper using a deep convolutional neural network.

Image-based deep learning (IBDL) is an advanced technique for predicting the surface irradiation conditions of laser surface processing technology. In pulsed-laser surface processing techniques, the number of superimposed laser shots is one of the fundamental and essential parameters that should be optimized for each material. Our primary research aims to build an adequate dataset using laser-irradiated surface images and to successfully predict the number of superimposed shots using the pre-trained deep convolutional neural network (CNN) models. First, the laser shot experiments were performed on copper targets using a nanosecond YAG laser with a wavelength of 532 nm. Then, the training data were obtained with the different superimposed shots of 1 to 1024 in powers of 2. After that, we used several pre-trained deep CNN models to predict the number of superimposed laser shots. Based on the dataset with 1936 images, VGG16 shows a high validation accuracy, higher sensitivity, and more than 99% precision than other deep CNN models. Utilizing the VGG16 model with high sensitivity could positively impact the industries' time, efficiency, and overall production.

An Efficient Approach to Face Emotion Recognition with Convolutional Neural Networks

Solutions for emotion recognition are becoming more popular every year, especially with the growth of computer vision. In this paper, classification of emotions is conducted based on images processed with convolutional neural networks (CNNs). Several models are proposed, both custom and transfer learning types. Furthermore, combinations of them as ensembles, alongside various methods of dataset modification, are presented. In the beginning, the models were tested on the original FER2013 dataset. Then, dataset filtering and augmentation were introduced, and the models were retrained accordingly. Two methods of emotion classification were examined: a multi-class classification, and a binary classification. In the former approach, the model returns the probability for each class. In the latter, separate models for each single class are prepared, together with an adequate dataset based on FER2013. Each model recognizes a single emotion from the others. The obtained results and a comparison of the applied methods across different models is presented and discussed.

Application and Analysis of Machine Learning Algorithms on Pima and Early Diabetes Datasets for Diabetes Prediction

Diabetes is a chronic condition that strike how your body burns food for energy. Much of the food you consume is converted by your body into sugar (glucose), which is then released into your bloodstream. Your pancreas releases insulin when your blood sugar levels rise. Over the years, several scholars have sought to create reliable diabetes prediction models. Due to a lack of adequate data sets and prediction techniques, this discipline still faces many unsolved research issues, which forces researchers to apply big data analytics and ML-based methodology. Four distinct machine learning algorithms are used in the study to analyze healthcare prediction analytics and solve the issues. In this investigation, the Pima and Early detection datasets were employed. We applied the Decision Tree, MLP, Naive Bayes, and Random Forest algorithms to these datasets and evaluated the accuracy and F-Measure. The goal of this research is to develop a system that could more precisely predict a patient's risk of developing diabetes.

English

In recent times, most of the Facial Recognition framework has shown better outcomes in a constrained environment. Due to its high usage, the FR has gained increased attention in the research field. However, it has faced various issues in the real-time application, such as facial disguise designing in facial features, temporal variations, and low-quality images. Therefore, there is a requirement for significant techniques for detecting and identifying the human face. This paper has implemented an effective FR technique. The adequate dataset relevant to the FR process is initially gathered from the standard dataset. Then, the images are pre-processed with the aid of the median filtering process to obtain the pre-processed images. Then, it is given to the three level-Discrete Wavelet Transform (DWT), which is used to attain the spectral features. Then, the features from the spectral features are optimally chosen by using the Improved Horse Herd Optimization algorithm termed (IHHO) derived from Horse Herd Optimization (HHO). Further, it is subjected to the Adaptive Cascaded framework termed ACLSDN, designed using Long Short Term Memory (LSTM) and Deep Neural Network (DNN) model. It is carried out by averaging the score attained from both the LSTM and DNN models for attaining the classified outcomes in maximized accuracy. Here, the same IHHO algorithm is used for optimizing the parameters in the LSTM and DNN frameworks. Finally, the effectiveness of the designed FR model is validated using various metrics and shows maximized accuracy value.

Unconstrained vocal pattern recognition algorithm based on attention mechanism

Deep learning-based voiceprint recognition methods rely heavily on adequate datasets, especially those closer to the natural environment and more complex under unconstrained conditions. Yet, the data types of open-source speech datasets are too homogeneous nowadays, and there are some differences with the address collected in natural application environments. For few Chinese datasets used, this paper proposes and produces an unconstrained Chinese speech dataset with richer data types closer to those collected in a natural environment. To address the inadequate extraction of acoustic features in the unconstrained speech dataset, a new two-dimensional convolutional residual network structure based on the attention mechanism is designed and applied to acoustic feature extraction. The residual block structure in the residual network is improved by the SE module and the CBAM module to obtain the SE-Cov2d and CSA-Cov2d models respectively. Finally, it is experimentally demonstrated that the attention mechanism can help the network focus on more critical feature information and fuse more differentiated features in feature extraction.

Wetlands estimation survey using DotS method: The case of Paya Indah Wetlands (PIW), Malaysia

Climate change poses a detrimental impact on the environment due to increasing carbon dioxide concentration in the atmosphere. In mitigating climate change, wetlands play a significant role in optimising the function of the earth's ecosystem for carbon sequestration and water resources rehabilitation. However, the effectiveness of wetlands management depends on an ecologically and statistically adequate dataset. Unfortunately, it is difficult for developing countries to implement area frame methods based on established sampling methods for data collection and monitoring. Thus, in this study, the dot sampling (DotS) method was adopted as it is simple, efficient, reliable and does not need a sample frame or high cost. DotS method uses the readily available platform, i.e., integration between Microsoft Excel-Macro (EM) and Google Earth (GE). This method systematically distributes the sample dots in GE's target area according to the locations generated by the EM. The field survey was then conducted at the sample dots of the target area in Paya Indah Wetlands (PIW) to ratify the site location. During the first stage, 72 dot sample locations were generated systematically, distributed across the target area, and 47 dot sample locations were located within the PIW. After the preparatory stage, 10 dot sample locations were eliminated. After the fieldwork inspection and evaluation, only 28 dot sample locations were selected as study stations. Hence, this study provides comprehensive data and examples applying the established DotS method with high reliability, accuracy, and rapidity in determining the number of study stations of wetlands. Based on the advantages, this study recommends the DotS method as a reliable survey method suitable for all study locations, especially in the ecology field. Applying the Dots methods as a surveying tool enables rapid land assessment in planning and suits the assessment of ex-mining constructed wetlands. K eywords: Climate change, constructed wetland, DotS method, estimation survey wetland, excel-macro, google earth

Developing precision agriculture using data augmentation framework for automatic identification of castor insect pests.

Castor (Ricinus communis L.) is an important nonedible industrial crop that produces oil, which is used in the production of medicines, lubricants, and other products. However, the quality and quantity of castor oil are critical factors that can be degraded by various insect pest attacks. The traditional method of identifying the correct category of pests required a significant amount of time and expertise. To solve this issue, automatic insect pest detection methods combined with precision agriculture can help farmers in providing adequate support for sustainable agriculture development. For accurate predictions, the recognition system requires a sufficient amount of data from a real-world situation, which is not always available. In this regard, data augmentation is a popular technique used for data enrichment. The research conducted in this investigation established an insect pest dataset of common castor pests. This paper proposes a hybrid manipulation-based approach for data augmentation to solve the issue of the lack of a suitable dataset for effective vision-based model training. The deep convolutional neural networks VGG16, VGG19, and ResNet50 are then adopted to analyze the effects of the proposed augmentation method. The prediction results show that the proposed method addresses the challenges associated with adequate dataset size and significantly improves overall performance when compared to previous methods.

Proposing Intelligent Approach to Predicting Air Kerma within Radiation Beams of Medical X-ray Imaging Systems.

The air kerma is a key parameter in medical diagnostic radiology. Radiologists use the air kerma parameter to evaluate organ doses and any associated patient hazards. The air kerma can be simply described as the deposited kinetic energy once a photon passes through the air, and it represents the intensity of the radiation beam. Due to the heel effect in the X-ray sources of medical imaging systems, the air kerma is not uniform within the X-ray beam's field of view. Additionally, the X-ray tube voltage can also affect this nonuniformity. In this investigation, an intelligent technique based on the radial basis function neural network (RBFNN) is presented to predict the air kerma at every point within the fields of view of the X-ray beams of medical diagnostic imaging systems based on discrete and limited measured data. First, a diagnostic imaging system was modeled with the help of the Monte Carlo N Particle X version (MCNPX) code. It should be noted that a tungsten target and beryllium window with a thickness of 1 mm (no extra filter was applied) were used for modeling the X-ray tube. Second, the air kerma was calculated at various discrete positions within the conical X-ray beam for tube voltages of 40 kV, 60 kV, 80 kV, 100 kV, 120 kV, and 140 kV (this range covers most medical X-ray imaging applications) to provide the adequate dataset for training the network. The X-ray tube voltage and location of each point at which the air kerma was calculated were used as the RBFNN inputs. The calculated air kerma was also assigned as the output. The trained RBFNN model was capable of estimating the air kerma at any random position within the X-ray beam's field of view for X-ray tube voltages within the range of medical diagnostic radiology (20-140 kV).

Color ghost imaging through a dynamic scattering medium based on deep learning

This paper presents a color computational ghost imaging scheme through a dynamic scattering medium based on deep learning that uses a sole single-pixel detector and is trained by a simulated data set. Due to the color distortion and noise sources being caused by the scattering medium and detector, a simulation data generation method is proposed accordingly that easily adapts to the actual environment. Adequate simulation data sets allow the trained artificial neural networks to exhibit strong reconfiguration capabilities for optical imaging results. It is worth noting that the network trained by our method can reconstruct better details of the image than the simulation data sets according to the ideal state. Its effectiveness is demonstrated in optical imaging experiments with both rotated double-sided frosted glass and a milk solution used as the dynamic scattering medium.

The impact of GDPR on data sharing for European cancer research

The European Groundshot—addressing Europe's cancer research challenges: a Lancet Oncology Commission by Mark Lawler and colleagues,1 which addresses Europe's cancer research challenges, has outlined how collaborative research is vital to ensure rapid improvement in cancer care from early diagnosis to precise application of therapies. Data sharing ensures research robustness, adequate datasets for statistically significant results, economies of scales and cost-effective approaches through data reuse and enhancement, and ultimately expedites the validation and implementation of research findings.

Evaluation of end-to-end aspect-based sentiment analysis methods employing novel benchmark dataset for aspect, and opinion review analysis

Aspect-based sentiment analysis (ABSA) deals with the determination of sentiments for opinion targets. While historically this research task has been addressed with pipeline approaches, more recent works use neural networks to jointly deal with the aspect term and opinion term extraction, as well as the polarity classification. Although learned together, most NN-based approaches and all pipeline approaches do not model correlations between the tasks. This is also based on the absence of adequate datasets which are annotated for all sub-tasks in a unified tagging scheme. We address this bottleneck and introduce the first purposely designed and annotated dataset for ABSA. The DAORA dataset covers 2,100 Tripadvisor reviews, and it is annotated on aspect terms, opinion terms, as well as aspect term polarity, using a unified tagging scheme. It was designed especially for end-to-end aspect-based sentiment analysis of real-world reviews and does not use any sentence repetition or removal. We evaluate the DAORA dataset in several experiments employing state-of-the-art models for ABSA. We set benchmarks and analyze the strengths as well as weaknesses of the data and approaches.

Gender biases in the training methods of affective computing: Redesign and validation of the Self-Assessment Manikin in measuring emotions via audiovisual clips.

Audiovisual communication is greatly contributing to the emerging research field of affective computing. The use of audiovisual stimuli within immersive virtual reality environments is providing very intense emotional reactions, which provoke spontaneous physical and physiological changes that can be assimilated into real responses. In order to ensure high-quality recognition, the artificial intelligence (AI) system must be trained with adequate data sets, including not only those gathered by smart sensors but also the tags related to the elicited emotion. Currently, there are very few techniques available for the labeling of emotions. Among them, the Self-Assessment Manikin (SAM) devised by Lang is one of the most popular. This study shows experimentally that the graphic proposal for the original SAM labelling system, as devised by Lang, is not neutral to gender and contains gender biases in its design and representation. Therefore, a new graphic design has been proposed and tested according to the guidelines of expert judges. The results of the experiment show an overall improvement in the labeling of emotions in the pleasure-arousal-dominance (PAD) affective space, particularly, for women. This research proves the relevance of applying the gender perspective in the validation of tools used throughout the years.

Comparative Analysis of Machine Learning Techniques for Predicting Air Pollution

The modern and motorized way of life has cultured air pollution. Air pollution has become the biggest rival of robust living. This situation is becoming more lethal in developing countries and so in Pakistan. Hence, this inquiry was carried out to propose an architecture design that could make real-time prediction of air pollution with another purpose of scanning the frequently adopted algorithm in past investigations. In addition, it was also intended to narrate the toxic effects of air pollution on human health. So, this research was carried out on a large dataset of Seoul as an adequate dataset of Pakistan was not attainable. The dataset consisted of three years (2017-2019) including 647,512 instances and 11 attributes. The four distinctive algorithms termed Random Forest, Linear Regression, Decision Tree and XGBoosting were employed. It was inferred that XGB is more promising and feasible in predicting concentration level of NO2, O3, SO2, PM10, PM2.5 and CO with the lowest RMSE and MAE values of 0.0111, 0.0262, 0.0168, 49.64, 41.68 and 0.1856 and 0.0067, 0.0096, 0.0017, 12.28, 7.63 and 0.0982 respectively. Furthermore, it was found out as well that the Random Forest was preferred mostly in the previous studies related to air pollution prophecy while many probes supported that air pollution is very detrimental to human health especially long-lasting exposure causes lung cancer, respiratory and cardiovascular diseases.