Dr. P. Priyadharshini and M. Vanitha Archana, Department of Mathematics, PSG College of Arts and Science, Coimbatore, India
An incompressible MHD nanofluid boundary layer flow over a stretching surface employing Buongiornos design investigated by considering the heat and mass transfer convective states. The Brownian motion and thermophoresis models were used to implement the nanofluid model. Operating the similarity transmutations, transforms the governing nonlinear partial differential equations to ordinary differential equations consisting of the multiple slips (momentum, energy, and concentration) and later worked by using a program written together with the stiffness shifting in Wolfram Language. The consequences of different physical parameters, magnetic parameter M, Brownian motion parameter Nb, thermophoresis parameter Nt, Lewis number Le, Suction parameter temperature and concentration Biot numbers on the multiple slips as well as the variation of the skin friction coefficient, reduced Nusselt and reduced Sherwood numbers concerning M and for various values of physical parameters obtained graphically, then compared with other recent works in tabular form. Finally, introduce a new environment to employ machine learning by performing the sensitivity analysis based on the iterative method for predicting the variation of skin friction coefficient, reduced Nusselt number and, reduced Sherwood number concerning the magnetic parameter. Here is investigated for estimating the cost function by comparing the actual and expected outcomes to fit the best fit line with the help of graphical forms. Machine learning gives a powerful and intelligent data processing architecture, and may hugely enrich the existing research techniques and industrial application of fluid mechanics.
Magnetohydrodynamics, Nanofluid, Stiffness Shifting, Machine Learning Algorithms.
Farhi Marir1, Hussein Fakhry1 and Aida Azar2, 1College of Technological Innovation, Zayed University, Academic City, Dubai, UAE, 2College of Medicine, Mohammed Bin Rashid University of Medicine and Health Sciences (MBRU), Dubai Health Care City, Dubai, United Arab Emirates
The advances in digital data collection and storage technology allowed the storage of huge amounts of medical publications in MEDLINE database which contains more than 25 million references to journal articles and abstracts in life sciences and biomedicine. This research work builds on Swanson use of mathematical association between A and C concepts/terms through a list of B concept/terms retrieved from medical articles that contain either A&B or B&C terms links A to C. Swanson discovered evidence that fish oil (A) cure vessel blood disorder (C) and magnesium (A) help with migraine headache (C) which were clinically approved two years later. In this research paper we present co-occurrence mining algorithm and an A&C pre-defined domain knowledge base (containing for instance Garlic Composition and Blood pressure causes) to filter and reduce the exponential number of shared B terms retrieved from MEDLINE articles using Swanson’s Arrowsmith machine. The reduced number of relevant B terms makes it easier to build scientific evidence to validate publicly known remedies for recurrent diseases like for instance garlic reduces blood pressure and honey relieves coughs and soothes a sore throat, etc.
Co-occurrence Text Mining, ABC Arrowsmith Discovery Machine, Dietary Aliments & Disease Knowledge Base, and MEDLINE medical database.
James Usher and Pierpaolo Dondio, School of Computing Technological University Dublin, Dublin, Ireland
After parliament failed to approve his revised version of the ‘Withdrawal Agreement’, UK Prime Minister Boris Johnson called a snap general election in October 2019 to capitalise on his growing support to ‘Get Brexit Done’. Johnson’s belief was that he had enough support countrywide to gain a majority to push his Brexit mandate through parliament based on a parliamentary seat majority strategy. The increased availability of large-scale Twitter data provides rich information for the study of constituency dynamics. In Twitter, the location of tweets can be identified by the GPS and the location field. This provides a mechanism for location-based sentiment analysis which is the use of natural language processing or machine learning algorithms to extract, identify, or distinguish the sentiment content of a tweet (in our case), according to the location of origin of said tweet. This paper examines location-based Twitter sentiment for UK constituencies per country and aims to understand if location-based Twitter sentiment majorities per UK constituencies could determine the outcome of the UK Brexit election. Tweets are gathered from the whisperings of the UK Brexit election on September 4th 2019 until polling day, 12th December 2019. A Naive Bayes classification algorithm is applied to assess political public Twitter sentiment. We identify the sentiment of Twitter users per constituency per country towards the political parties’ mandate on Brexit and plot our findings for visualisation. We compare the grouping of location-based sentiment per constituency for each of the four UK countries to the final Brexit election first party results per constituency to determine the accuracy of location-based sentiment in determining the Brexit election result. Our results indicate that location-based sentiment had the single biggest effect on constituency result predictions in Northern Ireland and Scotland and a marginal effect on Wales base constituencies whilst there was no significant prediction accuracy to England’s constituencies. Decision tree, neural network, and Naïve Bayes machine learning algorithms are then created to forecast the election results per constituency using location-based sentiment and constituency-based data from the UK electorate at national level. The predictive accuracy of the machine learning models was compared comprehensively to a computed-baseline model. The comparison results show that the machine learning models outperformed the baseline model predicting Brexit Election constituency results at national level showing an accuracy rate of 97.87%, 95.74 and 93.62% respectively. The results indicate that locationbased sentiment is a useful variable in predicting elections.
BREXIT Election, Twitter, Sentiment, UK Election.
Olawunmi George and Sheikh Iqbal Ahamed, Computer Science Department, Marquette University, Milwaukee, Wisconsin, USA
Motor imagery (MI) is arguably one of the most common brain-computer interface (BCI) paradigms. The decoding process, in many cases, involves the use of small amounts of data gathered over a period. The decoding performance might therefore be limited, due to the size of available data. Also, the non-stationarity of signals across sessions and subjects can pose a challenge to effective decoding. To solve these challenges, transfer learning is proposed as the suitable approach, which could yield optimal performance even with small amounts of data and handle the non-stationarity of signals with adaptation. It has been applied across domains and tasks where only small amounts of data are available and where signal distribution changes are more rapid. Transfer learning (TL) aids the decoding process by utilizing previous knowledge learnt in the decoding process to enhance future decoding. In this study, we apply the concept of transfer learning in the classifier space, yielding improvements in the decoding process up to 3%. We investigate its effect across two main scenarios: withinand across-subject motor imagery decoding. Within each of these scenarios, we consider how to optimally transfer useful knowledge, across sessions, to aid the decoding process, taking into consideration factors such as the update mode and data sources to be used for transfer across subjects. To our knowledge, this is one of the few works making such considerations in the application of transfer learning in electroencephalography (EEG)-based MI. From our results, we conclude that transfer learning is useful in motor imagery experiments, where small amounts of data are available, and can be used in mitigating the effect of non-stationarity across sessions and subjects.
EEG, BCI, motor imagery, deep learning, machine learning, transfer learning, knowledge transfer, supervised learning.
Yan Zhang, Xiangyang Feng and Ming Zhu, College of Computer Science and Technology, Donghua University, Shanghai, China
Facial expressions can express different emotions.Similar facial expressions usually correspond to same emotions, and the changing process of emotional states is reflected in the dynamic changes of facial expressions. However, existing studies mainly focus on instantaneous emotional states, which cannot reflect the intensity of emotions.This paper proposes a method to study the process of emotion change based on dynamic expression similarity, which can evaluate not only changes in emotional state but also changes in emotional intensity.First, the features of dynamic expressions are extracted based on the VGG16 network model. Then, the cosine similarity of the expression features is calculated to match the corresponding emotions. At the same time, the expression intensity of each frame is calculated to evaluate the change in emotional intensity. The experimental results show that the similarity calculated in this paper is increased by 9.7% on average, which can be used for the study of emotional state.
Dynamic Expression Similarity, The Emotional State Change, Emotional Intensity.
Glenda Tan Hui En1*, KoayTze Erhn1* and Shen Bingquan2, 1Raffles Institution, Singapore, 1Raffles Institution, Singapore, 2DSO National Laboratories, Singapore
More infectious virus variants can arise from rapid mutations in their proteins, creating new infection waves. These variants can evade one’s immune system and infect vaccinated individuals, lowering vaccine efficacy. Hence, to improve vaccine design, this project proposes Optimus PPIme– a deep learning approach to predict future, more infectious variants from an existing virus (exemplified by SARS-CoV-2). The approach comprises (i)an algorithm which acts as a “virus” attacking a host cell. To increase infectivity, the “virus” mutates to bind better to the host’s receptor. 2 algorithms were attempted –greedy searchand beam search. The strength of thisvariant-host bindingwas then assessed by a transformer network we developed, with a high accuracy of 90%. With both components, beam search eventually proposed more infectious variants. Therefore, this approach can potentially enable researchers to develop vaccines that provide protection against future infectious variants before theyemerge, pre-empting outbreaks and saving lives.
Virus Variants, Transformers, Deep Learning.
Ximeng Zhang1 and Yu Sun2, 1Yorba Linda High School, 19900 Bastanchury Rd, Yorba Linda, CA92886, 2California State Polytechnic University, Pomona, CA, 91768, Irvine, CA92620
Recycling is an essential measurement to change waste into reusable material. In the US, only about 30% of solid wastes are properly recycled, compared to other developed countries such as Northern Ireland (50.6%), Japan(50%), Schotland (46.9%), Wales (56.9%), it is much lower. However, in the US, the amount of solid waste disposal has increased in the past decade, which leads to air pollution, water pollution, soil pollution and solid waste is also a cause of many diseases. Specifically, it is noticed that many people have difficulty realizing how well they are doing in the process of recycling. Therefore, an app based on dart language is created to check how well people recycle through a scoring system and collect data from grocery receipts to see if the app can help make the consumers’ receipts full of more recyclable items. And it is hypothesized that this app can increase efficiency in recycling and promote people to encourage individuals to use more recyclable items. A clear trend in my data of scores gained from my family grocery receipt each week shows that the amount of recyclables increase as the weeks go by since the number from the app did increase. The number I get from recycling pops up in my head as I do the weekly grocery shopping with my family and reminds me to buy more recyclable items. The app is proven helpful and does increase recycling efficiency to 95%. The product, as an app, will be widely used by smartphones.
Environmental protection, Artificial Intelligence, NLP.
Anthony S. Deese1, Thomas Brennan2, Andrew Bechtel2, Joe Jesson1, and Efrain Rodriguez3, 1Department of Electrical and Computer Engineering, The College of New Jersey, Pennington, New Jersey, USA, 2Department of Civil Engineering, The College of New Jersey, Pennington, New Jersey, USA, 3Tenna LLC, Edison, New Jersey, USA
This paper discusses the design and testing of an application of digital IoT technology to track relatively small and inexpensive non-powered assets like concrete roadway barriers within a smart city, focusing on knowledge innovation, organizational systems and technology, and location intelligence. This application is of significant interest to many private corporations. The authors objective is to develop, implement, and test wireless digital IoT hardware having the ability to track construction assets while being sufficiently inexpensive to warrant their placement on smaller assets while having sufficient range to track and communicate with assets over ranges up to 2.5km. The proposed solution is durable enough to withstand the stresses of use on construction sites and sustainable enough to operate for 5+ years without manual intervention or introducing an external energy source to recharge the unit. This paper provides significant physical test results that validate the efficacy of the approach in practical use-case scenarios.
Internet of Things (IoT), Wireless Network, Transportation.
Christopher Harth-Kitzerow & Gonzalo Munilla Garrido, Technical University of Munich, Germany
An edge computing marketplace could enable IoT devices (Outsourcers) to outsource computation to any participating node (Contractors) in their proximity. In return, these nodes receive a reward for providing computation resources. In this work, we propose a scheme that verifies the integrity of arbitrary deterministic functions and is resistant to both dishonest Outsourcers and Contractors who try to maximize their expected payoff. We tested our verification scheme with state-of-the-art pre-trained Convolutional Neural Network models designed for object detection. On all devices, our verification scheme causes less than 1ms computational overhead and a negligible network bandwidth overhead of at most 84 bytes per frame. Our implementation can also perform our verification scheme’s tasks parallel to the object detection to eliminate any latency overhead. Compared to other proposed verification schemes, our scheme resists a comprehensive set of protocol violations without sacrificing performance.
Edge Computing, Internet of Things, Function Verification, Computing Marketplaces.
Oluwadare Oatunji1, Oluwashola Adeniji2 and Ukeme Sunday3, 1Department of Computer Science, University of Ibadan, Nigeria, 2Department of Computer Science, University of Ibadan, Nigeria, 3Lagos Business School, Lagos, Nigeria
Internet of things often referred to as (IOT) refers to the collective network of connected devices and the technology that facilitates communication between devices and the cloud. It comprises various components that enhances high level of technology in the digital world. IOT has helped to revolutionize the world in various fields such as digital tech, E-Health, Artificial Intelligence, Cybernetics, E-currency, robotic science and various aspect of lives. Since the introduction of IOT in the New Millennium, it has risen to be the back bone Organization via its Network. Despite the enormous benefit of IOT, security and privacy challenges has been a major factor militating against it. This is feasible via: insecurities, zero-day attacks, cyber threats, cyberattacks etc. Internet hijackers and cyber fraud stars are threatening the future of IOT.
cyber-attack, Internet of things (IOT), Network Telescope, Computational Algorithm, security, Hardware, software, cloud computing.
Rawabi Nazal Alhamad1 and Faeiz M. Alserhani2, 1Department of Computer Sciences, Jouf University, Saudi Arabia, 2Department of Computer Engineering and Networks, Jouf University, Saudi Arabia
The rapid use of the Internet of Things (IoT) has influenced many domains including smart cities, cameras, wearables, smart industrial equipment, and other aspects of our daily lives. The IoT environment, on the other hand, deals with a huge data that needs to be kept secure from tampering or theft. This study proposes five deep and machine learning techniques for identifying malware in network traffic based on the IoT-23 dataset. Random Forest, XGBoost, Convolutional Neural Network (CNN), Catboost, and Long Short-Term Memory (LSTM) models are among the classifiers utilized. The dataset was preprocessed to remove unnecessary or missing data, and then the most significant features were extracted using a feature engineering technique. The highest overall accuracy achieved was 96% by applying all classifiers except LSTM which recorded a lower accuracy.
Internet of things (IoT), Malware Detection, Feature Selection, Random Forest, Catboost, CNN, LSTM, XGBoost.
Tahani M. Alshammari1 and Faeiz M. Alserhani2, 1Department of Computer Sciences, Jouf University, Saudi Arabia, 2Department of Computer Engineering and Networks, Jouf University, Saudi Arabia
The adoption of the Internet of Things (IoT) is gaining traction due to the rapid development in smart devices and decreased prices of sensing devices. These IoT devices are more vulnerable to being attacked or compromised. In this research, we propose an SDN-based IoT Anomaly intrusion detection system which detects abnormal behaviors and attacks as early as possible. Three machine learning (ML) techniques, that is support vector machines (SVM), k-nearest neighbor (KNN) and random forest (RF) are used to design an IDS which is intended to be placed at the SDN controller in order to watch and learn the behavior of IoT devices over time, with any deviation from normal behavior being labeled as an attack. On the selected benchmark dataset, we tested and evaluated our model. We reported a comparison of the outcomes of the three machine learning algorithms, which show that Random Forest (RF) algorithm outperformed support vector machines (SVM) and k-nearest neighbor (KNN).
Internet of things (IoT), intrusion detection systems (IDS), Machine learning (ML), Software Defined networks (SDN).
Charith Jayakantha, Charm Theekshana, Malith Gayashan, Isuru Kumarasiri, Dhammika De Silva, Dinuka Wijendra, Faculty of Computing, Sri Lanka Institute of Information Technology (SLIIT), Malabe, Sri Lanka
Sri Lanka faces floods almost in every year. The main reason for floods in Sri Lanka is, the lack of a proper system for predicting rainfall and floods and informing the public. As a solution, this research proposes to create a proper weather forecasting and water level forecasting system and visualize the flooding areas in a google map. In this research LSTM Model, Facebook prophet, grey model, Convolutional Neural Network and Arima model are used as forecasting models. In addition, create a device, which connected with IoT sensors to retrieve the real-time weather and flood water levels. Then real-time data are monitored in a web application. Then, after executing all predictions and validations which is implemented, highlight the spots which can be affected by flood disaster. And finally implement a chatbot to provide extensive knowledge about real-time weather data and forecasted data and others.
Internet of things (IoT), Arima Model, Facebook Prophet, Grey Model, Convolutional Neural Network.
Samantha Hilario, Nicole Mallari, Martin Dizon, Inigo Salazar, Joel Mendoza, University of Asia and the Pacific, Pasig City, Metro Manila, Philippines
The researchers discovered that Company A is experiencing challenges in managing and tracking their furniture, appliances, and IT resources. The company currently uses MS Excel as their asset management; however, it became inefficient enough to be used in their business operations anymore since the company is growing and gaining more clients. The researchers meet with the client to discuss the issues, business processes, possible solutions, and other constraints. The researchers implemented an Agile Methodology approach in developing the Resource Allocation Management System [RAMS]. In addition, they would provide a web-based system that keeps track of the whole asset’s life cycle (item’s specifications, procurement, current user and location, and usage history) where every asset record is protected. In conclusion, the researchers found that the Agile approach was effective in identifying issues and designing solutions that tailor-fit to the client’s business.
Asset management, Agile, Asset, Resources, MS Excel.
M.A. Ba Humaish, School of Science and Engineering, The American University in Cairo, Cairo, Egypt
Routing the packets from the source to the destination is the main responsibility of the network layer, so many routing algorithms in this layer can be used to select the best routes and data structure of packets. In this paper, we have used two routing protocols Routing Information Protocol (RIP) and Open Shortest Path First (OSPF) to analyze the performance of a network. Different applications such as video, FTP, print, and voice were configured based on different parameters. Additionally, we have looked at the same networks behavior when we implement Quality of service using Weighted Fair Queuing algorithm (WFQ) and Multi-Protocol Label Switching protocol (MPLS) to ensure that the multimedia networks performance is a guaranteed endeavor. The simulation results analyzed using the Optimized Network Engineering Tools (OPNET), and we compared these protocols to the effectiveness and performance in the network topology, also Multimedia applications are analyzed also, and the most essential QoS requirements are determined and quantified.
Quality of Service (QoS), Multimedia Applications, Open Shortest Path First (OSPF), Routing Information Protocol (RIP), Weighted Fair Queuing algorithm (WFQ) and Multi-Protocol Label Switching (MPLS).
Lul Mohamed Osman1 and DR. COLLINS ODUOR2, 1Student, Department of Computing, United States International University-Africa, Nairobi, Kenya, 2Assistant Professor, School of Science and Technology, United States International University-Africa, Nairobi, Kenya
The study aimed to identify the usability challenges of E-banking and mobile banking applications. Both quantitative and qualitative analyses were used in this study. For qualitative analysis, a review and analysis were done. This enabled the researcher to collect information on identifying the usability challenges of E-banking and mobile banking applications. This studys findings show that the users of internet banking and mobile banking applications encountered several challenges such as language barrier due to the use of English and Arabic languages in the banking applications. Some banking applications could not adapt to the different screen sizes. Compatibility issues were noted hence users on various mobile device operating systems such as iOS encountered issues accessing the applications. Some applications page take more time to load, and also, response time to customers is low. Some applications dont have a support center and no user guide to problems and application usage.
Usability, E-banking, mobile banking.
Evgeny Volynsky, Merlin Mehmed and Stephan Krusche, Technical University of Munich, Munich, Germany
The migration from a monolithic to a microservice architecture is a recurring step in many software projects. In this paper, we describe a case study of a migration process of the learning management system Artemis. We migrate the shared database into multiple databases based on the database-per-microservice pattern. We developed Architect, a framework which is based on a domain-specific language for building dependable distributed systems as a template to ensure the data consistency of the distributed transactions using the Saga pattern. The Architect framework helped to reduce the complexity of using the Saga pattern. It introduced the eventual consistency in a distributed database system and decreased the coupling of the data storage. Architect currently does not provide a software evolution approach. We will add support for reengineering projects, which can facilitate the migration process of existing system.
Microservices architecture, Migration of monolithic application, Saga pattern, Domain-specific-language, Distributed transactions.
Mahmood Radhi Hadi1 and Adnan Saher Mohammed2, 1Department of Computer Engineering, Karabük University, Karabük, Turkey, 2Adnan Saher Mohammed, Karabük University, Karabük, Turkey
Software-Defined Networking (SDN) is the next generation to change the architecture of traditional networks. SDN is one of the promising solutions to change the architecture of internet networks. Attacks become more common due to the centralized nature of SDN architecture. It is vital to provide security for the SDN. In this study, we propose a Network Intrusion Detection System-Deep Learning module (NIDS-DL) approach in the context of SDN. Our suggested method combines Network Intrusion Detection Systems (NIDS) with many types of deep learning algorithms. Our approach employs 12 features extracted from 41 features in the NSL-KDD dataset using a feature selection method. We employed classifiers (CNN, DNN, RNN, LSTM, and GRU). When we compare classifier scores, our technique produced accuracy results of (98.63%, 98.53%, 98.13%, 98.04%, and 97.78%) respectively. Our proposed approach was successful in binary classification and detecting attacks, implying that our approach (NIDS-DL) might be used with great efficiency in the future.
Network Intrusion Detection System, Software Defined Networking, Deep Learning.
Fabrizio dAmore, Department of Computer, Control and Management Engineering, Sapienza University of Rome, Italy
In this paper we propose an approach to the design of processes and software that aims at decreasing human and software errors, that so frequently happen, making affected people using and wasting a lot of time for the need of fixing the errors. We base our statements on the natural relationship between quality and error rate, increasing the latter as the error rate decreases. We try to classify errors into several types and address techniques to reduce the likelihood of making mistakes, depending on the type of error. We focus on this approach related to organization, management and software design that will allow to be more effective and efficient in this period where mankind has been affected by a severe pandemic and where we need to be more efficient and effective in all processes, aiming at an industrial renaissance which we know to be not too far and easily reachable once the path to follow has been characterized, also in the light of the experience.
Errors, Quality, Processes, Governance, Digitalization, Digital hygiene, Information technology, Computer Science.
Jazzar Nesrine1 and Ali Douik2, 1University of Sfax, National Enginneering School of Sfax, Noccs Lab, Tunis, Tunisia, 2National Enginneering School of Sousse, University of Sousse
Ischemic stroke, brain cells death due to a lack of oxygen, is a leading cause of long-term disability and death. Accurate diagnosis and timely intervention can effectively improve the blood supply of the ischemic stroke area and minimize brain damage. Recent studies have shown the potential to use magnetic resonance imaging (MRI) to provide contrast imaging to visualize and detect lesions. However, manual segmentation of the stroke lesion produced by MRI is a tedious and timeconsuming task. Therefore, the automatic ischemic stroke lesion segmentation method may shows excellent advantages. In this paper, we propose a novel deep learning method used to detect and localize brain ischemic stroke, a generalization encoder-decoder by modifying U-Net architecture. We integrate multi-path architecture into both encoder and decoder blocks to captures different levels of the encoded state, which helps in more robust decision-making for stroke lesion segmentation. In bottleneck of the architecture, we applied dilated blocks to improve the underlying predictive capabilities. The proposed method has been tested on the publicly accessible web platform provided by the MICCAI Ischemic Stroke Lesion Segmentation (ISLES) challenge. The results demonstrate that the proposed method achieves a mean dice coefficient 0.91 of with the training and 0.84 with the testing data respectively.
Ischemic stroke segmentation, Convolutional neural network, UNet, MRI, Dilated blocks.
Farida Mustafazade1 and Peter F. Ebbinghaus2, 1GAM Systematic, 2Teufel Audio
We propose a cross-encoder augmented BERTScore model for semantic answer similarity trained on our new dataset of non-common names in English contexts. There are several issues with the existing general machine translation or natural language generation evaluation metrics, and question answering (QA) systems are indifferent in that sense. To build robust QA systems, we need the ability to have equivalently robust evaluation systems to verify whether model predictions to questions are similar to ground-truth annotations. The ability to compare similarity based on semantics as opposed to pure lexical overlap is important to compare models fairly and to indicate more realistic acceptance criteria in real-life applications. We build upon the first to our knowledge paper that uses transformer-based model metrics to assess semantic answer similarity and achieve higher correlations to human judgement in the case of no lexical overlap.
Question-answering, semantic textual similarity, exact match, pre-trained language models, cross-encoder, bi-encoder.
Abderrahim Bouderbane, Department of English, University of Mila, Algeria
This research is conducted to investigate the characteristics of speaking fluently and its components like: pauses, filled pauses, and hesitations. These components are very important in speaking, and they are used to generate ideas, plan what to say next and organize the content. the aim of this research is to divide samples of recorded speech extracts of non-native speakers of English to distinguish between pauses, fillers and hesitations. A sample of ten native speakers (British), and 10 non native speakers (Algerian teachers of english) are evaluated using a software called Praat. This software is mainly used in phonetics to describe pitch, stress and intonation. This software can also be used to calculate the total time of speaking, and demonstrate the pauses graphically as they are produced by speakers.The measurement of one pause is made reliable by Praat as far as the measurement is adjustable at all levels (Styler 2013). The adjustment can start from a 0.00Hz to 5000 or 6000 Hz. This software cannot tell where one word start and where it ends. As such, the researcher needs to segment sound filles with information when using any sort of automated information. This is generally done by creating a textgrid annotation in a text file, which is saved separetly. These annotations are composed of different tiers which are marked either by intervals or specific points in the file. The results found that pauses and hesitations are more frequent and longer in non-native speaking production.
Fluency, speech processing, native speakers, non-native speakers, Pratt software.
Louangtrath, P.I, Bangkok University International, Thailand
The purpose of this paper is to determine the rate of reproduction of COVID-19 in Thailand during the delta strain outbreak between June and August 2021. In so doing, the ASEAN countries were used as cohort group for comparison study. The data used in this research was obtained from the corona tracker website. The compartmental model was used for determining the rate of reproduction (Ro) where the population N is divided into three compartments. The rate of reproduction can be used as an indicator for COVID-19 spreading. If it means that the disease has died out. If it means that the disease is expanding. Using the generalized logistic function, the initial value was Q = 3,440; K = 18,912 and A = 2,290. We determined the growth rate for Thailand to be thus B = 8.5. By allowing C =1 we can solve for v with the end value of on July 31, 2021. By solving for v we obtain v = 0.05. With known v, we can find the solution to the generalized logistic equation of COVID-19 for Thailand’s daily infection. Our finding shows that the lag phase for Thailand is . We used series for the four types of COVID-19 variants to generate a linear equation under Weibull’s system analysis. We obtained a range between and. When X = 1 the range is making the herd immunity for Thailand to fall within a range of . These findings are useful to policy makers, public health officials and the general public because stakeholders can better be informed and take necessary preventive measures against the spread of COVID-19.
Compartmental Model, COVID-19, exponential function, growth model, Herd immunity, logistic function.
Mohamed Sofiane BATTA1, 2, Zibouda ALIOUAT2, Hakim MABED1, and Malha MERAH2, 1FEMTO-ST Institute/DISC, University of Bourgogne Franche-Comte, Montbeliard, France, 2LRSD Laboratory, Computer Science Dept, Ferhat Abbas University Setif 1, Setif, Algeria
Energy efficiency is a major concern and a critical issue for energy constrained wireless networks. In this context, clustering is commonly used for topology man- agement and maximizing the network lifetime. Clustering approaches typically use a multi-hopping mechanism where Cluster Heads (CHs) near the Base Sta- tion (BS) consume higher energy since they relay data of farther CHs. Therefore, nodes close to the BS are strangled with an overloaded routing task and tend to die earlier than their intended lifetime, which affects the network performance. This situation is known as the hot spot problem that induces unbalanced energy consumption among CHs. The concern in this work is to address the intra- clustering structure in large scale environments to tolerate the network scaling and reasonably balance the energy consumption among CHs. In this regard, we propose a new Unequal Clustering algorithm based on Kruskal heuristic (UCKA) to optimize the network lifetime. UCKA applies the Kruskal heuristic in a dis- tributed fashion to perform a minimum spanning tree within large cluster which strengthen the intra-cluster routing structure and reduce the energy devoted to wireless communications. To the best of our knowledge, this is the first solution that combines the Kruskal heuristic and the unequal clustering to extend the devices durability and alleviate the hot spot problem. Simulation results indi- cate that UCKA can effectively reduce the energy consumption and lengthen the network lifetime.
IoT, WSN, Energy-aware protocols, Unequal Clustering, Hot spot energy prob- lem, Kruskal Heuristic.
Ponthip Limlahapun and Puntip Jongkroy, Geography Department, Kasetsart University, Bangkok, Thailand
His Majesty King Bhumibol Adulyadej of Thailand, conferred with the title King Bhumibol or King Rama IX, was the ninth monarch of Thailand in the Chakri dynasty. His Majesty initiated more than 2,000 projects throughout his 70 year reign. He worked with diligence and inspired the Thai people in their living, gratitude, and morality. The Royal Development Projects Board (RDPB) collects all information, approves plans and activities, and directs, monitors, and coordinates the operation of government agencies and state enterprises concerning Royal Development Projects. The various kinds of projects spread all over the country, from north to south. The geolocation would present and benefit younger generations who may not realize where those projects took place. More than 50 years ago, his Majesty made an effort to visit the entirety of Thailand, even inconvenient places with poor road systems. Therefore, this research aims to develop an online web spatial database of the Royal Projects under the inspiration of King Rama 9th, which will manage various information (i.e., project name, covered areas, responsibility agencies) systematically based on geographic locations. Online data could be accessed anywhere, anytime, and at any place, and users can select their interest area/project based on places and project types, which is more flexible compared to information in printed text formats. The research process was comprised of data requested from the office of the RDPB, organizing it in a format used for further geospatial web development, as well as locating the projects’ location and designing a map display. The process discovered that the geocoding function embedded in a geographic information system (GIS) software (i.e., ArcGIS or QGIS) could not identify the locations in Thailand; therefore, written programming was used to search for the location in Google, which worked faster. Since there are approximately more than 2000 projects, online mapping should consider how quantitative data relates in a spatial form with map communication, aesthetics, and sufficient information. GIS could be applied in a variety of fields i.e., disaster, public facilities, utilities, etc. Data presentations in only number or text format may provide information that lacks project distribution or referenced surrounding areas. Therefore, the interactive map would help supplement the existing information to make it clearer and easier to understand. Additional, the virtual learning platform was developed to attract young generations.
geolocation, geospatial web-based system, royal projects, virtual learning platform.
Enrique Hern´andez-Orallo1, Antonio Armero2, Carlos Tavares Calafate1 and Pietro Manzoni1, 1Department of Computer Engineering (DISCA), Universitat Polit`ecnica de Val`encia, Valencia, Spain, 2Department of Engineering Projects, Polit`ecnica de Val`encia, Valencia, Spain
Tracker tags, such as Apple’s AirTags and Tile’s trackers, are small and cheap devices aimed at finding lost things. Tags use Bluetooth technology and community tracking to find the location of lost tags whenever a compatible smartphone passes nearby. When we lose a tag, several questions may arise: how are the chances of finding it? Or, when will we receive a notification of its detection? Unfortunately, companies such as Apple or Tile do not provide technical information about the efficiency of their tags. Thus, this paper aims to give answers to those questions based on a methodological analysis and evaluation of tags’ efficiency. The main aspects that impact the efficiency of finding a lost tag are the detection range, community size, and human mobility. These aspects are evaluated in a campus scenario, showing that the probability of finding lost tags is very high: more than 99% of them were found, and the average detection time was around 1 hour.
Mobile Crowdsensing Opportunistic networks Internet of Things Performance Evaluation.
G. N. R. Prasad1 and B. RAMADASU2, 1Senior Assistant Professor, Dept. of Master of Computer Applications,Chaitanya Bharathi Institute of Technology (Autonomous), Gandipet, Hyderabad – 500 075, India, 2Assistant Professor, Dept. of Computer Science and Engineering,Chaitanya Bharathi Institute of Technology (Autonomous), Gandipet, Hyderabad – 500 075, India
Blockchainis the latest buzzwords in the field of Computer Science since the past four years. It has the potential to completely take over the traditional Financial Institutions we know since the past century and bring about Web 3.0. In order to do this the technology has to be highly secure. To understand how and why blockchain is so secure and reliable we need to first understand how the different blocks are linked together in a Blockchain Network. The blocks in a blockchain are linked with the help of a Hashing Algorithm. A hashing algorithm is a one-way mathematical function which generates a unique output for every unique input. The Bitcoin Network uses a SHA-256 hash. A basic block has a data, hash and previous hash field. The hash value of any block is dependent on the hash value of the previous block, therefore making it very difficult to alter the data as any tampering would require altering hashes of each and every subsequent block in the blockchain. A typical blockchain network can have millions of blocks on its chain. This makes it virtually impossible to alter the blockchain, thereby making it so secure. Blockchain is a decentralized technology . It has a lot of power when it comes to solving business difficulties.Algorithms on the nodes validate blockchain transactions. A transaction cannot be initiated by a single entity. Finally, blockchains ensure transparency by allowing each participant to view transactions at any moment. Smart contracts ensure safe transactions, reducing the risk of third-party interruption. Decentralization and immutability are two key characteristics of blockchain. Faster transactions, transaction and validation in seconds, and so forth.
blockchain, cryptocurrency, hash, blocks.
Amit Kumar Dwivedi1, Naveen Kumar1 and Manik Lal Das2, 1IIIT Vadodara, Gujarat, India, 2DA-IICT, Gandhinagar, Gujarat, India
Data outsourcing in a third-party storage server is a costeffective business model, which is widely used by small and medium-sized enterprises. As data is managed by a third party, the accountability of the involved entities becomes an important aspect of this business model. Proof of violation of agreed terms between the data owner, users, and the storage service provider ensures accountability and is desirable as it enhances the trust of the data owner towards other system entities. Furthermore, strong accountability identifies the misbehaving entity’s activities and protects the honest entity from false accusations. In this paper, a strong, accountable distributed auditing scheme is proposed for outsourcing data in a public cloud setup. A proof-based solution with fine-grained read-write access is used to address the strong accountability and auditing of data as well as services. The scheme is analyzed and compared with the related schemes. The experimental results of the scheme show that the proposed scheme is efficient and practical.
Accountability, Distributed audit, Data integrity, Distributed cloud.
Tahani M. Alshammari, Department of Computer and information sciences, Aljouf University, Saudi Arabia
In this paper, investigated the accuracy of classification of two free online deepfake detection tools. The results indicate the tools tested achieve very high performance when applied to datasets in which training and test data shares same cameras, types of manipulation and processing pipelines, but once applied to a black box dataset performance degrade. Most of the current models and methods developed for deepfake detecting depend heavily on the alignment between training set and test data. This demands taking a broader view and considering solutions that go beyond image and video analysis.
DeepFake Detection, Deep learning, Datasets, Deepware.
Jenny Sun1 and Yu Sun2, 1JSerra Catholic High school, 26351 Junipero Serra Rd, San Juan Capistrano, CA92675, 2California State Polytechnic University, Pomona, CA, 91768, Irvine, CA92620
Most teenagers have some types of skin blemishes and allergies that often go undiagnosed . This can leadtomany unnecessary skin ailments and treatments that do not address the root of the problems . This is especiallytrue for young women who are beginning to use makeup. This can lead to embarrassments and even isolation and depression. But with this Intelligent Mobile Platform app, teenagers, parents, and dermatologists can all be assistedin identifying specific skin conditions such as oily or dry skin, skin prone to breakouts and rashes, acne, or evenmore severe conditions such as eczema and psoriasis . Once this app has made a proper diagnosis, it will recommend and prescribe proper skin treatments intended to avoid or solve potentially major physical andemotional problems. This paper is designed to discuss the ways that this app evolved into a productive andreputable device capable of improving the lives of millions of young teenagers as well as those of older ages whocontinue to struggle with skin problems.
Natural Language Processing, Artificial Intelligence, Mobile Platform.
Mohamed Ahmed M. Khalifa1, Hebat-Allah M. Mourad2 and Mahmoud Abdelaziz3, 1Electronics and Electrical Communications Dept, Faculty of Engineering, Cairo University, Giza, Egypt, 2Electronics and Electrical Communications Dept, Faculty of Engineering, Cairo University, Giza, Egypt, 3Communications and Information Engineering, University of Science and Technology in Zewail City, Giza, Egypt
Increasing global data traffic made 5G (IMT-2020) a good solution, especially in the mid-band spectrum (such as the 3.5 GHz range) because it offers a good balance between coverage and capacity. Thus, some countries had identified it as one of the candidate bands for 5G. C-band (from 3400 to 4200 MHz) is a mainstay of satellite communications and is used to provide broadband connectivity in remote areas today. This paper discusses the feasibility of 5G (IMT-2020) and Fixed Satellite Service (FSS) system to coexist in the C-band range by analyzing the impact of the interference from 5G (IMT-2020) base stations towards the FSS earth station. This analysis is based on the most recent unwanted emissions limits used from 3GPP TS 38.104. The results show that in the adjacent channel scenario and by employing an elevation angle of 48° and a guard band from 41-100 MHz, 5G (IMT-2020) base station needs to be separated at least 0.295 Km away from the FSS earth station. The protection distance increases by 26.35 Km when decreasing the guard band to 30 MHz. So, a new unwanted emission limit is proposed to reduce the protection distance in the 30 MHz guard band scenario.
FSS, interference, 5G, C-band, IMT-2020 & 5G BS.
Lahcene Mekadem and Malika Bourenane, Faculty of Exact & Applied Sciences, University of Oran 1, Oran, Algeria
Faced with the evolution of mobile wireless networks (MANET) and the diversity of its fields of application, it has become necessary to develop a more precise intrusion detection system by exploiting information extracted from several layers of the OSI model. This article proposes a cross layer intrusion detection system to mitigate black hole attacks in mobile wireless networks. The proposed system consists of two levels, at the first level the data extracted from several layers are used as inputs to a fuzzy logic system to determine the nature of a particular node (malicious node, normal node or suspicious node).At the second level, if a node is suspect, Dempster- Shafers theory will be applied to investigate that node. The proposed system, evaluated using the NS2 simulator, is found to be effective in detecting and isolating the malicious node.
Manet, Cross layer, Intrusion detection system, Fuzzy logic, Dempster-Shafers theory.
Copyright © ArIT 2022