Search

Due to advances in technology, social media has become the most popular medium for spreading news. Many messages are published on social media sites such as Facebook, Twitter, Instagram, etc. Social media platforms also provide opportunities to express opinions and social phenomena such as hate, offensive language, racism, sexual content, and all forms of verbal violence, which have amazingly increased. These behaviors do not only affect specific countries, groups, or societies but extend beyond these areas into people's daily lives. This study examines sexual content and harassment discourse in Arabic social media to build an accurate system for detecting sexual harassment expressions. The dataset was collected from Twitter posts to make the classification. A deep learning model was developed as a classification system to identify sexual speech using Bidirectional Long-Short-Term Memory (BiLSTM), Temporal Convolutional Network (TCN) with word embedding and the FastText previously trained on the Arabic language model. The proposed (TCN-BiLSTM) model was compared with Extreme Gradient Boosting (XGBoost). The CASH dataset implemented with the (TCN -Bi-LSTM) model gate obtained an accuracy rate of 96.65% and an F0.5 value of 0.969. The implementation of XGBoost using word embeddings resulted in an accuracy rate of 92.56% and an F0.5 value of 0.925. Findings and manual interpretation showed that different text representation methods with various deep learning algorithms obtain higher classification performance easily in complex sentences. This strategy is helpful with languages that are difficult to study morphologically, like Arabic, Turkish, and Lithuanian.

In conjunction with the global concern regarding the spread of fake news on social media, there is a large flow of research to address this phenomenon. The wide growth in social media and online forums has made it easy for legitimate news to merge with comprehensive misleading news, negatively affecting people’s perceptions and misleading them. As such, this study aims to use deep learning, pre-trained models, and machine learning to predict Arabic and English fake news based on three public and available datasets: the Fake-or-Real dataset, the AraNews dataset, and the Sentimental LIAR dataset. Based on GloVe (Global Vectors) and FastText pre-trained models, A hybrid network has been proposed to improve the prediction of fake news. In this proposed network, CNN (Convolution Neural Network) was used to identify the most important features. In contrast, BiGRU (Bidirectional Gated Recurrent Unit) was used to measure the long-term dependency of sequences. Finally, multi-layer perceptron (MLP) is applied to classify the article news as fake or real. On the other hand, an Improved Random Forest Model is built based on the embedding values extracted from BERT (Bidirectional Encoder Representations from Transformers) pre-trained model and the relevant speaker-based features. These relevant features are identified by a fuzzy model based on feature selection methods. Accuracy was used as a measure of the quality of our proposed models, whereby the prediction accuracy reached 0.9935, 0.9473, and 0.7481 for the Fake-or-Real dataset, AraNews dataset, and Sentimental LAIR dataset respectively. The proposed models showed a significant improvement in the accuracy of predicting Arabic and English fake news compared to previous studies that used the same datasets.

Recently, the phenomenon of the spread of fake news or misinformation in most fields has taken on a wide resonance in societies. Combating this phenomenon and detecting misleading information manually is rather boring, takes a long time, and impractical. It is therefore necessary to rely on the fields of artificial intelligence to solve this problem. As such, this study aims to use deep learning techniques to detect Arabic fake news based on Arabic dataset called the AraNews dataset. This dataset contains news articles covering multiple fields such as politics, economy, culture, sports and others. A Hybrid Deep Neural Network has been proposed to improve accuracy. This network focuses on the properties of both the Text-Convolution Neural Network (Text-CNN) and Long Short-Term Memory (LSTM) architecture to produce efficient hybrid model. Text-CNN is used to identify the relevant features, whereas the LSTM is applied to deal with the long-term dependency of sequence. The results showed that when trained individually, the proposed model outperformed both the Text-CNN and the LSTM. Accuracy was used as a measure of model quality, whereby the accuracy of the Hybrid Deep Neural Network is (0.914), while the accuracy of both Text-CNN and LSTM is (0.859) and (0.878), respectively. Moreover, the results of our proposed model are better compared to previous work that used the same dataset (AraNews dataset).

Semantic segmentation is an exciting research topic in medical image analysis because it aims to detect objects in medical images. In recent years, approaches based on deep learning have shown a more reliable performance than traditional approaches in medical image segmentation. The U-Net network is one of the most successful end-to-end convolutional neural networks (CNNs) presented for medical image segmentation. This paper proposes a multiscale Residual Dilated convolution neural network (MSRD-UNet) based on U-Net. MSRD-UNet replaced the traditional convolution block with a novel deeper block that fuses multi-layer features using dilated and residual convolution. In addition, the squeeze and execution attention mechanism (SE) and the skip connections are redesigned to give a more reliable fusion of features. MSRD-UNet allows aggregation of contextual information, and the network goes without needing to increase the number of parameters or required floating-point operations (FLOPS). The proposed model was evaluated on three multimodal datasets: polyp, skin lesion, and nuclei segmentation. The obtained results proved that the MSDR-Unet model outperforms several state-of-the-art U-Net-based methods.

Junko Iwatsuki,1,2 Takeshi Kondo,3 Noriyuki Takahashi,3– 5 Hideki Takami,3 Hiroshi Nishigori,6 Itzel Bustos-Villalobos,7 Branko Aleksic,7,8 Hideki Kasuya,7 Nobutaro Ban,9 Tetsuya Yagi,10 Norbert Skokauskas11,12 1International Collaboration Planning Center, Institute of International Education and Exchange, Nagoya University, Nagoya, Japan; 2Present Affiliation is Office for International Education and Exchange, Faculty and Graduate School of Economics, Kobe University, Kobe, Japan; 3Center for Postgraduate Clinical Training and Career Development, Nagoya University Hospital, Nagoya, Japan; 4Department of General Medicine/Family & Community Medicine, Nagoya University Graduate School of Medicine, Nagoya, Japan; 5Department of Education for Community-Oriented Medicine, Nagoya University Graduate School of Medicine, Nagoya, Japan; 6Center for Medical Education, Nagoya University Graduate School of Medicine, Nagoya, Japan; 7Department of International Medical Education, Nagoya University Graduate School of Medicine, Nagoya, Japan; 8Department of Psychiatry, Nagoya University Graduate School of Medicine, Nagoya, Japan; 9Medical Education Center, Aichi Medical University School of Medicine, Nagakute, Japan; 10Department of Infectious Diseases, Nagoya University Graduate School of Medicine, Nagoya, Japan; 11Regional Centre for Child and Youth Mental Health and Child Welfare, Department of Mental Health, Norwegian University of Science and Technology, Trondheim, Norway; 12Department of Child and Adolescent Psychiatry, St. Olavs University Hospital, Trondheim, NorwayCorrespondence: Branko AleksicNagoya University Graduate School of Medicine, 65 Tsurumai-Cho, Showa-ku, Nagoya, 466-8550, JapanTel +81 52 7442282Fax +81 52 7442293Email branko@med.nagoya-u.ac.jpPurpose: Japanese higher education institutions have long been striving for the globalization of medical education. Nagoya University (NU) adopted PBL as a means of enhancing intercultural awareness in globalizing medical education by working with the Norwegian University of Science and Technology (NTNU), Faculty of Medicine and Health Science, under the Trondheim NTNU-Nagoya (TroNa) partnership for mobility and internationalization of child and mental health studies. This study aims to assess students’ attitudes towards PBL and to suggest future developments in this form of education by introducing common PBL scenarios experienced at NTNU and NU.Methods: Two 90-minute PBL sessions were conducted at NU. Ten groups of medical students were formed, each consisting of up to 10 students, and students were asked to fill in a questionnaire developed to assess their understanding of, attitudes to and satisfaction with the classes. We investigated three different groups of questions on: NU medical students’ general impressions of PBL; their impressions of PBL in child and adolescent psychiatry (CAP); and their impressions of PBL in specific case scenarios. Correlations between each of the questions from the three groups were evaluated using multivariate analysis.Results: Overall, a majority of the NU medical students were satisfied with PBL, while a small number preferred traditional lecture-style learning (5%). More than half of the students agreed that PBL increased their understanding and interest in CAP (53%), although some male students felt that the amount of time spent was insufficient (20.3%). Correlations were seen for students who thought that PBL enhanced their understanding of and interest in CAP. Regarding case scenarios, most students (82.5%) agreed that PBL helped them to develop clinical problem-solving skills.Conclusion: The study found an overall positive attitude towards PBL, PBL in CAP and the specific PBL case scenario presented.Keywords: problem-based learning, child psychiatry, children and mental health, English medical education, internationalization

Cyberbullying is a new form of online violence that has emerged as a result of the social media industry's explosive expansion since it allows for indirect communication. Despite the usefulness of digital media, it has been used by weak people to threaten and bully others online. In the last ten years, research has shown that children and teenagers are increasingly experiencing cyberbullying as a concern. This paper examines the research conducted during the previous ten years, categorizes it, and presents statistical analyses of the data collected during that time. A table is used to present various data, including the dataset that was used, its size (number of samples, posts, or messages), the methods that were employed, and the metrics that were gathered from the examined research that was taken from the publications that were looked into. This survey will be helpful to everyone who wants to advance their understanding of how machine learning may be used to identify cyberbullying, and it may help create a social media environment that is safe and relatively healthy by automatically identifying bullying communications. [ABSTRACT FROM AUTHOR]

Social media and news agencies are major sources for tracking news and events. With these sources' massive amounts of data, it is easy to spread false or misleading information. Given the great dangers of fake news to societies, previous studies have given great attention to detecting it and limiting its impact. As such, this work aims to use modern deep learning techniques to detect Arabic fake news. In the proposed system, the attention model is adapted with bidirectional long-short-term memory (Bi-LSTM) to identify the most informative words in the sentence. Then, a multilayer perceptron (MLP) is applied to classify news articles as fake or real. The experiments are conducted on a newly launched Arabic dataset called the Arabic Fake News Dataset (AFND). The AFDN dataset contains exactly 606912 news articles collected from multiple sources, so it is suitable for deep learning requirements. Both simple recurrent neural networks (S-RNN), long short-term memory (LSTM), and gated recurrent units (GRU) are used for comparison. According to evaluation criteria, our proposed model achieved an accuracy of (0.8127), which is the best and highest accuracy among the deep learning methods used in this work. Moreover, the performance of our proposed model is better compared to previous studies, which used the AFND. [ABSTRACT FROM AUTHOR]

Cyberbullying is one of the biggest electronic problems that takes multiple forms of harassment using various social media. Currently, this phenomenon has become very common and is increasing, especially for young people and adolescents. Negative comments have a significant and dangerous impact on society in general and on adolescents in particular. Therefore, one of the most successful prevention methods is to detect and block harmful messages and comments. In this research, negative Arabic comments that refer to cyberbullying will be detected using a support vector machine algorithm. The term frequency-inverse document frequency vectorizer and the count vectorizer methods were used for feature extraction, and the results were improved using the cuckoo search algorithm. The resulting accuracy before and after optimizing the support vector machine’s hyperparameters is 85.8% and 87.1%, respectively. [ABSTRACT FROM AUTHOR]

Daiji Nagayama,1,2 Haruki Imamura,2 Yuta Sato,2 Takashi Yamaguchi,2 Noriko Ban,2 Hidetoshi Kawana,2 Masahiro Ohira,2 Atsuhito Saiki,2 Kohji Shirai,3 Ichiro Tatsuno2 1Center of Endocrinology and Metabolism, Shin-Oyama City Hospital, Oyama-City, 2Center of Diabetes, Endocrinology and Metabolism, 3Department of Vascular Function, Sakura Medical Center, Toho University, Sakura-City, Japan Objective: The objective of this study is to investigate the association of body mass index (BMI) with arterial stiffness assessed by cardioankle vascular index (CAVI). Subjects and methods: A retrospective cross-sectional study was conducted in 23,257 healthy Japanese subjects (12,729 men and 10,528 women, aged 47.1 ± 12.5 years, BMI 22.9 ± 3.4 kg/m2) who underwent health screening between 2004 and 2006 in Japan. Exclusion criteria were current medication use and a past history of cardiovascular disease, hypertension, stroke, diabetes, and nephritis. Results: Male subjects showed significantly higher BMI, CAVI, and triglycerides and lower high-density lipoprotein (HDL)-cholesterol compared with female subjects. Next, the subjects were divided into tertiles of BMI: lower, middle, and upper, in a gender-specific manner. After adjusting for confounders including age, systolic blood pressure, and HDL-cholesterol identified by multiple regression analysis, the mean CAVI decreased progressively as BMI tertile increased in both genders. Furthermore, a negative inverse relationship between BMI and adjusted CAVI was observed throughout the BMI distribution. Multivariate logistic regression model for contributors of high CAVI (≥90th percentile) identified obesity (odds ratios (95% confidence interval): 0.804 (0.720–0.899)], older age [15.6 (14.0–17.4)], male gender [2.26 (2.03–2.51)], hypertension [2.28 (2.06–2.54)], impaired fasting glucose [1.17 (1.01–1.37)], and low HDL-cholesterol [0.843 (0.669–1.06)] as independent factors. Conclusion: We demonstrated an inverse relationship between CAVI and BMI in healthy Japanese subjects, suggesting that systemic accumulation of adipose tissue per se may lead to a linear decrease of arterial stiffness in nonobese and obese subjects without metabolic disorders. Keywords: BMI, cardioankle vascular index, arterial stiffness

Highly purified virions of satellite tobacco mosaic virus (STMV) were found to crystallize at relatively low concentrations (300-500 microg ml(-1)) in pure water. Small crystals of these preparations were examined in the transmission electron microscope after either being rotary shadowcast with metal or negatively stained with 4% uranyl acetate. Stereo views were also obtained of both types of preparations. Stereo pairs of metal-coated crystals provided good three-dimensional images. When stereo pairs of negatively stained crystals were printed from second negatives, they provided striking images although the three-dimensional aspect was not so pronounced. Images of both types of preparations were compared with a computer-generated model of the virus. This model was based on data obtained in earlier X-ray diffraction crystallographic studies. Measurements of crystal axes on the EM images were somewhat lower than those of the computer model. It is assumed the reason for this is the dehydration of crystals during preparation for electron microscopy. The EM images did verify the type of crystal lattice determined in the X-ray diffraction studies. Conversely, knowing the exact unit cell parameters and the distribution of virions in the crystal from X-ray diffraction data aids in the further interpretation of electron micrographs of virus crystals.

The number of users of social media sites has increased nowadays, and while these sites have many benefits, they also have many damages that have grown with the increasing number of users. Among these damages that have spread in social media sites in our time is the phenomenon of cyberbullying. It has become necessary to find solutions to detect it to prevent and hold bullies accountable to reduce the phenomenon of cyberbullying, which has great health and mental effects on the victim in society. There have been many attempts to build models to detect and classify cyberbullying by using machine learning and deep learning algorithms with different sets of data that were collected from social media sites such as Twitter, YouTube, Facebook, Instagram, and others. In this work, we show a group of previous studies that used machine learning and deep learning algorithms in good attempts to detect and classify the phenomenon of cyberbullying. [ABSTRACT FROM AUTHOR]

Detailed structures are now available for three plant satellite viruses, satellite tobacco necrosis virus (STNV), satellite tobacco mosaic virus (STMV), and satellite panicum mosaic virus (SPMV). It is, therefore, possible to compare the tertiary structure of viral protein subunits, their quaternary interactions, and the interactions of protein subunits with the RNA genome. This analysis indicates that, in spite of common function and preservation of a "jelly-roll" motif in the protein monomer, the three viruses are remarkably different. The differences include the arrangement of secondary structural elements, interactions of adjacent subunits, and the disposition of subunits relative to icosahedral symmetry axes. In each of the three viruses, however, the narrow end of the jelly roll forms fivefold contacts. The fivefold protein interactions are organized about a Ca2+ ion for STNV, an anion for STMV, and, apparently, neither of these for SPMV. Low-resolution neutron diffraction studies using H2O/D2O solvent contrast variation revealed the general location of the RNA genome within the STNV. In the case of SPMV, regions of electron density on the interior of the capsid could be assigned to RNA, although it was not possible to model the nucleic acid. Only for STMV was nucleic acid visible in election density maps, and this was manifested as double-helical RNA segments associated with each coat protein dimer. The observations presented here provide no support for any common evolutionary relationship.

Satellite panicum mosaic virus (SPMV) has been purified from pearl millet and obtained in a variety of different crystal forms, at least two of which appear suitable for high resolution X-ray diffraction analysis. The first is of cubic space group P4(2)32 with a = b = c = 183.1 A and two virus particles in the unit cell. The second is of a primitive orthorhombic space group with a = 166.1 A, b = 266.7 A and c = 269.1 A, with four virus particles in the unit cell. While the cubic crystal has as its asymmetric unit one twelfth of the icosahedron, or five capsid protein subunits, the asymmetric unit of the orthorhombic crystals is an entire particle. The cubic crystals diffract to at least 2.8 A resolution. We have also succeeded in crystallizing, but not yet characterizing the master virus, PMV.