Journal of Supercomputing最新文献

Based on the supernetwork theory, a two-step rumor detection model was proposed. The first step was the classification of users on the basis of user-based features. In the second step, non-user-based features, including psychology-based features, content-based features, and parts of supernetwork-based features, were used to detect rumors posted by different types of users. Four machine learning methods, namely, Naive Bayes, Neural Network, Support Vector Machine, and Logistic Regression, were applied to train the classifier. Four real cases and several assessment metrics were employed to verify the effectiveness of the proposed model. Performance of the model regarding early rumor detection was also evaluated by separating the datasets according to the posting time of posts. Results showed that this model exhibited better performance in rumor detection compared to five benchmark models, mainly owing to the application of the supernetwork theory and the two-step mechanism.

As the number of users getting acquainted with the Internet is escalating rapidly, there is more user-generated content on the web. Comprehending hidden opinions, sentiments, and emotions in emails, tweets, reviews, and comments is a challenge and equally crucial for social media monitoring, brand monitoring, customer services, and market research. Sentiment analysis determines the emotional tone behind a series of words may essentially be used to understand the attitude, opinions, and emotions of users. We propose a novel long short-term memory (LSTM)-convolutional neural networks (CNN)-grid search-based deep neural network model for sentiment analysis. The study considers baseline algorithms like convolutional neural networks, K-nearest neighbor, LSTM, neural networks, LSTM-CNN, and CNN-LSTM which have been evaluated using accuracy, precision, sensitivity, specificity, and F-1 score, on multiple datasets. Our results show that the proposed model based on hyperparameter optimization outperforms other baseline models with an overall accuracy greater than 96%.

Gigantic growth of social media and unbeatable trend of progress in the direction of the web seeking user's interests have generated a storm of social text streams. Seeking information to know the phenomenon of various events in the early stages is quite interesting. Various kinds of social media live streams attract users to participate in real-time events to become a part of an immense crowd. However, the vast amount of text is present on social media, the unnecessary information bogs a social text stream filtering to extract the appropriate topics and events effectively. Therefore, detecting, classifying, and identifying burst events is quite challenging due to the sparse and noisy text of Twitter. The researchers' significant open challenges are the effective cleaning and profound representation of the text stream data. This research article's main contribution is to provide a detailed study and explore bursty event detection in the social text stream. Thus, this work's main motive is to present a concise approach that classifies and detects the event keywords and maintains the record of the event based on related features. These features permit the approach to successfully determine the booming pattern of events scrupulously at different time span. Experiments are conducted and compared with the state-of-the-art methods, which reveals that the proposed approach is proficient to detect valuable patterns of interest and also achieve better scoresto extract burst events on social media posted by various users.

Reliability analysis is one of the crucial issues for any scalable optical interconnection network. Torus is a highly scalable optical interconnect for data centre networks. The traditional torus network has XY routing algorithm. We have proposed a novel optimised routing algorithm. This paper focuses on the time-dependent and time-independent analysis for both terminal and broadcast reliabilities of the torus network using XY and optimised routing algorithm under various network sizes ( $N\times N$ where $N=8,16,32,64$ ). The results are evaluated and compared considering nodes failures in MATLAB.

Machine learning algorithms are becoming more and more useful in many fields of science, including many areas where computational methods are rarely used. High-performance Computing (HPC) is the most powerful solution to get the best results using these algorithms. HPC requires various skills to use. Acquiring this knowledge might be intimidating and take a long time for a researcher with small or no background in information and communications technologies (ICTs), even if the benefits of such knowledge is evident for the researcher. In this work, we aim to assess how a specific method of introducing HPC to such researchers enables them to start using HPC. We gave talks to two groups of non-ICT researchers that introduced basic concepts focusing on the necessary practical steps needed to use HPC on a specific cluster. We also offered hands-on trainings for one of the groups which aimed to guide participants through the first steps of using HPC. Participants filled out questionnaires partly based on Kirkpatrick's training evaluation model before and after the talk, and after the hands-on training. We found that the talk increased participants' self-reported likelihood of using HPC in their future research, but this was not significant for the group where participation was voluntary. On the contrary, very few researchers participated in the hands-on training, and for these participants neither the talk, nor the hands-on training changed their self-reported likelihood of using HPC in their future research. We argue that our findings show that academia and researchers would benefit from an environment that not only expects researchers to train themselves, but provides structural support for acquiring new skills.

In every field of life, advanced technology has become a rapid outcome, particularly in the medical field. The recent epidemic of the coronavirus disease 2019 (COVID-19) has promptly become outbreaks to identify early action from suspected cases at the primary stage over the risk prediction. It is overbearing to progress a control system that will locate the coronavirus. At present, the confirmation of COVID-19 infection by the ideal standard test of reverse transcription-polymerase chain reaction (rRT-PCR) by the extension of RNA viral, although it presents identified from deficiencies of long reversal time to generate results in 2-4 h of corona with a necessity of certified laboratories. In this proposed system, a machine learning (ML) algorithm is used to classify the textual clinical report into four classes by using the textual data mining method. The algorithm of the ensemble ML classifier has performed feature extraction using the advanced techniques of term frequency-inverse document frequency (TF/IDF) which is an effective information retrieval technique from the corona dataset. Humans get infected by coronaviruses in three ways: first, mild respiratory disease which is globally pandemic, and human coronaviruses are caused by HCoV-NL63, HCoV-OC43, HCoV-HKU1, and HCoV-229E; second, the zoonotic Middle East respiratory syndrome coronavirus (MERS-CoV); and finally, higher case casualty rate defined as severe acute respiratory syndrome coronavirus (SARS-CoV). By using the machine learning techniques, the three-way COVID-19 stages are classified by the extraction of the feature using the data retrieval process. The TF/IDF is used to measure and evaluate statistically the text data mining of COVID-19 patient's record list for classification and prediction of the coronavirus. This study established the feasibility of techniques to analyze blood tests and machine learning as an alternative to rRT-PCR for detecting the category of COVID-19-positive patients.