Pub Date : 2020-10-28DOI: 10.1109/UEMCON51285.2020.9298050
Giovanni Perin, David Nophut, L. Badia, F. Fitzek
Wireless broadcast routing is a complex problem, shown in the literature to be NP-complete. Current protocols implement either heuristics to find solutions that are not guaranteed to be optimal or classic flooding. However, many future use cases, like automotive applications, industrial robotics, and multimedia broadcast, will require efficient yet reliable methods. In this work, we use contextual multi-armed bandits together with opportunistic routing (OR) and network coding (NC) to find approximately optimal solutions to the problem of broadcast routing in a distributed fashion. Each router independently learns its own transmission credit, i.e., the number of packets to forward for each innovative packet received, so that the airtime cost, subject to latency constraints, is minimized. Results show that the proposed solutions, particularly the deep learning based one, vastly improve the overall reliability, while performing close to MORE multicast in terms of airtime and to B.A.T.M.A.N. in latency, both being the best candidates in the respective discipline among the tested ones.
{"title":"Maximizing Airtime Efficiency for Reliable Broadcast Streams in WMNs with Multi-Armed Bandits","authors":"Giovanni Perin, David Nophut, L. Badia, F. Fitzek","doi":"10.1109/UEMCON51285.2020.9298050","DOIUrl":"https://doi.org/10.1109/UEMCON51285.2020.9298050","url":null,"abstract":"Wireless broadcast routing is a complex problem, shown in the literature to be NP-complete. Current protocols implement either heuristics to find solutions that are not guaranteed to be optimal or classic flooding. However, many future use cases, like automotive applications, industrial robotics, and multimedia broadcast, will require efficient yet reliable methods. In this work, we use contextual multi-armed bandits together with opportunistic routing (OR) and network coding (NC) to find approximately optimal solutions to the problem of broadcast routing in a distributed fashion. Each router independently learns its own transmission credit, i.e., the number of packets to forward for each innovative packet received, so that the airtime cost, subject to latency constraints, is minimized. Results show that the proposed solutions, particularly the deep learning based one, vastly improve the overall reliability, while performing close to MORE multicast in terms of airtime and to B.A.T.M.A.N. in latency, both being the best candidates in the respective discipline among the tested ones.","PeriodicalId":433609,"journal":{"name":"2020 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117165035","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-10-28DOI: 10.1109/UEMCON51285.2020.9298034
Zihan Xu, Li Jiang, Yiduo Liang, Yong Jia, G. Cui, Longfei Tan
MIMO through-wall-radar is traditionally utilized to perform forward-view imaging (FVI) for the hidden targets behind the wall. Due to the limited view angle and target occlusion, FVI usually suffers from the problem of target missing in the case of multiple targets. In this paper, a novel down-view imaging (DVI) mode is presented to obtain robust three-dimensional (3D) multiple target images without target missing. Specifically, the propagation characteristic is analyzed for the 3D DVI in an enclosed building space. Then the DVI algorithm is introduced. Specifically, the back-projection algorithm is applied to form a 3D image and the exponential phase coherence factor (EPCF) weighting is adopted to suppress the multi-path ghosts. Based on the gprMax simulation results, in the presence of target occlusion, it is demonstrated that the presented DVI method has the ability to implement robust imaging for multiple targets, while the FVI misses the obscured target.
{"title":"Three-Dimensional Down-View Imaging Based on MIMO Through-Wall-Radar","authors":"Zihan Xu, Li Jiang, Yiduo Liang, Yong Jia, G. Cui, Longfei Tan","doi":"10.1109/UEMCON51285.2020.9298034","DOIUrl":"https://doi.org/10.1109/UEMCON51285.2020.9298034","url":null,"abstract":"MIMO through-wall-radar is traditionally utilized to perform forward-view imaging (FVI) for the hidden targets behind the wall. Due to the limited view angle and target occlusion, FVI usually suffers from the problem of target missing in the case of multiple targets. In this paper, a novel down-view imaging (DVI) mode is presented to obtain robust three-dimensional (3D) multiple target images without target missing. Specifically, the propagation characteristic is analyzed for the 3D DVI in an enclosed building space. Then the DVI algorithm is introduced. Specifically, the back-projection algorithm is applied to form a 3D image and the exponential phase coherence factor (EPCF) weighting is adopted to suppress the multi-path ghosts. Based on the gprMax simulation results, in the presence of target occlusion, it is demonstrated that the presented DVI method has the ability to implement robust imaging for multiple targets, while the FVI misses the obscured target.","PeriodicalId":433609,"journal":{"name":"2020 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON)","volume":"105 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127150121","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-10-28DOI: 10.1109/UEMCON51285.2020.9298175
P. Sharma, Damian Valles
Recognition technology has gained state of art performance with the dawn of deep convolutional neural network and with these achievements in the field of computer vision, machine learning and 3D sensor, industries are near to start new era of the automation. However, object detection for robotic grasping in varying environment, low illumination, occlusion and partial images gives poor accuracy and speed to detect object. In this research, a multimodal architecture is designed to be used as a base network/ backbone network of Single Shot Detector (SSD). This architecture uses RGB and Depth images as an input and gives single output. Most of the researchers used VGG16/19, ResNet and MobileNet for detection purposes. In this paper, a new architecture is designed to perform a specific task of grasping. For classification using RGB-D architecture, it achieved an average accuracy of 95% with the learning rate of 0.0001 and outperforms the other architectures in accuracy for limited objects.
{"title":"Backbone Neural Network Design of Single Shot Detector from RGB-D Images for Object Detection","authors":"P. Sharma, Damian Valles","doi":"10.1109/UEMCON51285.2020.9298175","DOIUrl":"https://doi.org/10.1109/UEMCON51285.2020.9298175","url":null,"abstract":"Recognition technology has gained state of art performance with the dawn of deep convolutional neural network and with these achievements in the field of computer vision, machine learning and 3D sensor, industries are near to start new era of the automation. However, object detection for robotic grasping in varying environment, low illumination, occlusion and partial images gives poor accuracy and speed to detect object. In this research, a multimodal architecture is designed to be used as a base network/ backbone network of Single Shot Detector (SSD). This architecture uses RGB and Depth images as an input and gives single output. Most of the researchers used VGG16/19, ResNet and MobileNet for detection purposes. In this paper, a new architecture is designed to perform a specific task of grasping. For classification using RGB-D architecture, it achieved an average accuracy of 95% with the learning rate of 0.0001 and outperforms the other architectures in accuracy for limited objects.","PeriodicalId":433609,"journal":{"name":"2020 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON)","volume":"150 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123226820","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-10-28DOI: 10.1109/UEMCON51285.2020.9298045
Narayana Darapaneni, D. Reddy, A. Paduri, Pooja Acharya, H. S. Nithin
The recent outbreak of COVID-19 in different states of India has major concerns for all administrative departments of the government and general public. The Pandemic has been tested positive in 1287945 individuals with 817209 recovered and 30601 succumbed to the disease. The first case of the novel coronavirus was detected in India on 30 January 2020. There was a lockdown imposed by the Government of India from 24 March 2020 and ended on 31 May 2020. A forecast in no lockdown scenario would help us to track the further progress of the disease and make sufficient data available in order to plan the future of hospital facilities, pharmaceutical investment etc.
{"title":"Forecasting of COVID-19 in India Using ARIMA Model","authors":"Narayana Darapaneni, D. Reddy, A. Paduri, Pooja Acharya, H. S. Nithin","doi":"10.1109/UEMCON51285.2020.9298045","DOIUrl":"https://doi.org/10.1109/UEMCON51285.2020.9298045","url":null,"abstract":"The recent outbreak of COVID-19 in different states of India has major concerns for all administrative departments of the government and general public. The Pandemic has been tested positive in 1287945 individuals with 817209 recovered and 30601 succumbed to the disease. The first case of the novel coronavirus was detected in India on 30 January 2020. There was a lockdown imposed by the Government of India from 24 March 2020 and ended on 31 May 2020. A forecast in no lockdown scenario would help us to track the further progress of the disease and make sufficient data available in order to plan the future of hospital facilities, pharmaceutical investment etc.","PeriodicalId":433609,"journal":{"name":"2020 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123767710","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-10-28DOI: 10.1109/UEMCON51285.2020.9298130
Shaimaa Ezz-ElDin, Omar Nabil, Hussam Murad, Farah Adel, Ahmed AbdEl-Jalil, K. Salah, Ayub Khan
In this paper, a python-implemented infrastructure of a CNN-based multi-object detector in autonomous driving using the single shot detector (SSD) is presented. The infrastructure consists of both training and inference for object detection. The main contribution of this paper is the design of the default anchor boxes tiling that reduce the amount of computations by simplifying the software implementation of the SSD object detector. This simplification is done by reducing the data path of the proposed detector. Moreover, a decrease in the inference time of the detector is the result of using tiled defaults boxes and a small number of layers in the VGG CNN. In addition, the CNN model presents an advantage in terms of high confidence boxes prediction. The proposed approach is faster due to the reduced number of layers and computations. The segmentation design of the input image anchor boxes is introduced to explain the software implementation. In addition, both the training and validation loss variations along the period of the training are illustrated.
{"title":"MINI-SSD: A Fast Object Detection Framework in Autonomous Driving","authors":"Shaimaa Ezz-ElDin, Omar Nabil, Hussam Murad, Farah Adel, Ahmed AbdEl-Jalil, K. Salah, Ayub Khan","doi":"10.1109/UEMCON51285.2020.9298130","DOIUrl":"https://doi.org/10.1109/UEMCON51285.2020.9298130","url":null,"abstract":"In this paper, a python-implemented infrastructure of a CNN-based multi-object detector in autonomous driving using the single shot detector (SSD) is presented. The infrastructure consists of both training and inference for object detection. The main contribution of this paper is the design of the default anchor boxes tiling that reduce the amount of computations by simplifying the software implementation of the SSD object detector. This simplification is done by reducing the data path of the proposed detector. Moreover, a decrease in the inference time of the detector is the result of using tiled defaults boxes and a small number of layers in the VGG CNN. In addition, the CNN model presents an advantage in terms of high confidence boxes prediction. The proposed approach is faster due to the reduced number of layers and computations. The segmentation design of the input image anchor boxes is introduced to explain the software implementation. In addition, both the training and validation loss variations along the period of the training are illustrated.","PeriodicalId":433609,"journal":{"name":"2020 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125360727","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-10-28DOI: 10.1109/UEMCON51285.2020.9298134
Faisal Alghayadh, D. Debnath
The network system of smart homes using a Internet of Things (IoT) device is increasing in parallel with cybersecurity challenges as these loT devices have some vulnerabilities such as hardware and software limitations that leads to difficulties with time to fit security features to any IoT systems. Therefore, the Intrusion Detection Systems (IDS) is the suggested method to mitigate these cyberattacks and monitor the requests in smart homes. IDS has the capacity to protect the smart home network and detect real-time vulnerabilities and threats. In this paper, we applied and compared four types of machine learning algorithms which are random forest, xgboost, decision tree, and k-nearest neighbors on two sorts of datasets. We randomly selected three samples from each dataset. The results show that our models for each algorithm can effectively achieve a satisfying seemingly classification accuracy with the lowest false positive rate.
{"title":"Performance Evaluation of Machine Learning for Prediction of Network Traffic in a Smart Home","authors":"Faisal Alghayadh, D. Debnath","doi":"10.1109/UEMCON51285.2020.9298134","DOIUrl":"https://doi.org/10.1109/UEMCON51285.2020.9298134","url":null,"abstract":"The network system of smart homes using a Internet of Things (IoT) device is increasing in parallel with cybersecurity challenges as these loT devices have some vulnerabilities such as hardware and software limitations that leads to difficulties with time to fit security features to any IoT systems. Therefore, the Intrusion Detection Systems (IDS) is the suggested method to mitigate these cyberattacks and monitor the requests in smart homes. IDS has the capacity to protect the smart home network and detect real-time vulnerabilities and threats. In this paper, we applied and compared four types of machine learning algorithms which are random forest, xgboost, decision tree, and k-nearest neighbors on two sorts of datasets. We randomly selected three samples from each dataset. The results show that our models for each algorithm can effectively achieve a satisfying seemingly classification accuracy with the lowest false positive rate.","PeriodicalId":433609,"journal":{"name":"2020 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115040704","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-10-28DOI: 10.1109/UEMCON51285.2020.9298185
Devaj Parikh, K. George
A link can be established between the human brain and an external device utilizing the Brain-Computer Interface technique which uses Electroencephalogram (EEG) signals. We can reduce car accidents occurring due to short-braking by applying this technique to the brakes for an automobile. This paper presents a system based on signals from the cerebellum part of the brain to control the brakes of an automobile. The system comprises of an ultra-cortex headset, personal computers with Processing IDE, and an Arduino board to control the braking mechanism. Three subjects tested the system where each subject performed four trials. Testing was performed to determine the time difference between the system to complete the action and the human to perform the same. The average time response measured was found to be 450ms for a human and 250ms for the system.
{"title":"Conceptual Neuroadaptive Brain Computer Interface for Autonomous Control of Automobile Brakes","authors":"Devaj Parikh, K. George","doi":"10.1109/UEMCON51285.2020.9298185","DOIUrl":"https://doi.org/10.1109/UEMCON51285.2020.9298185","url":null,"abstract":"A link can be established between the human brain and an external device utilizing the Brain-Computer Interface technique which uses Electroencephalogram (EEG) signals. We can reduce car accidents occurring due to short-braking by applying this technique to the brakes for an automobile. This paper presents a system based on signals from the cerebellum part of the brain to control the brakes of an automobile. The system comprises of an ultra-cortex headset, personal computers with Processing IDE, and an Arduino board to control the braking mechanism. Three subjects tested the system where each subject performed four trials. Testing was performed to determine the time difference between the system to complete the action and the human to perform the same. The average time response measured was found to be 450ms for a human and 250ms for the system.","PeriodicalId":433609,"journal":{"name":"2020 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON)","volume":"478 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116188042","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-10-28DOI: 10.1109/UEMCON51285.2020.9298132
Jordan Juliano, Jaron Lin, Alex Erdogan, K. George
Identification of pulse radar signals is a crucial component in radar receiver processing. Extracting RF radar pulse information in wideband receivers in a dense and noisy environment without prior knowledge of the signal is challenging in high-risk real-time scenarios. Overlapping radar pulses can create complications in identification by creating interference between the signals, causing loss or hidden information. This paper presents a deinterleaving overlapping radar pulse train process by correlating the pulse descriptor words (PDW) of a power envelope based deinterleaving algorithm with a parallel fast Fourier transform (FFT) based deinterleaving algorithm implemented on an MPSoC FPGA.
{"title":"Radar Pulse on Pulse Identification Parallel FFT and Power Envelope Algorithm","authors":"Jordan Juliano, Jaron Lin, Alex Erdogan, K. George","doi":"10.1109/UEMCON51285.2020.9298132","DOIUrl":"https://doi.org/10.1109/UEMCON51285.2020.9298132","url":null,"abstract":"Identification of pulse radar signals is a crucial component in radar receiver processing. Extracting RF radar pulse information in wideband receivers in a dense and noisy environment without prior knowledge of the signal is challenging in high-risk real-time scenarios. Overlapping radar pulses can create complications in identification by creating interference between the signals, causing loss or hidden information. This paper presents a deinterleaving overlapping radar pulse train process by correlating the pulse descriptor words (PDW) of a power envelope based deinterleaving algorithm with a parallel fast Fourier transform (FFT) based deinterleaving algorithm implemented on an MPSoC FPGA.","PeriodicalId":433609,"journal":{"name":"2020 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON)","volume":"240 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124644924","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-10-28DOI: 10.1109/UEMCON51285.2020.9298104
Abdulrahman Aldhaheri, Je Lee, Khaled Almgren
There are many events that occur in e-commerce platforms, which can be used to detect and understand the behavior of online users. Behavior analyses of e-commerce users can be utilized to impact both customers and businesses. Behavior analysis seeks to find useful information from clickstreams, which can be used to address challenging problems. Clickstreams quantify users’ movements based on the items they click on an e-commerce website. This work aims to mine clickstreams to predict users’ genders. The proposed approach utilizes deep learning and has been tested on a real-world dataset; the proposed approach outperformed others in terms of accuracy.
{"title":"An Event Detection Platform to Detect Gender Using Deep Learning","authors":"Abdulrahman Aldhaheri, Je Lee, Khaled Almgren","doi":"10.1109/UEMCON51285.2020.9298104","DOIUrl":"https://doi.org/10.1109/UEMCON51285.2020.9298104","url":null,"abstract":"There are many events that occur in e-commerce platforms, which can be used to detect and understand the behavior of online users. Behavior analyses of e-commerce users can be utilized to impact both customers and businesses. Behavior analysis seeks to find useful information from clickstreams, which can be used to address challenging problems. Clickstreams quantify users’ movements based on the items they click on an e-commerce website. This work aims to mine clickstreams to predict users’ genders. The proposed approach utilizes deep learning and has been tested on a real-world dataset; the proposed approach outperformed others in terms of accuracy.","PeriodicalId":433609,"journal":{"name":"2020 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON)","volume":"98 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130419468","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-10-28DOI: 10.1109/UEMCON51285.2020.9298158
Maryam Heidari, James H. Jones
Millions of online posts about different topics and products are shared on popular social media platforms. One use of this content is to provide crowd-sourced information about a specific topic, event, or product. However, this use raises an important question: what percentage of the information available through these services is trustworthy? In particular, might some of this information be generated by a machine, i.e., a "bot" instead of a human? Bots can be, and often are, purposely designed to generate enough volume to skew an apparent trend or position on a topic, yet the consumer of such content cannot easily distinguish a bot post from a human post. This paper introduces a new model that uses Bidirectional Encoder Representations from Transformers (Google Bert) for sentiment classification of tweets to identify topic-independent features for the social media bot detection model. Using a Natural Language Processing approach to derive topic-independent features for the new bot detection model distinguishes this work from previous bot detection models. We achieve 94% accuracy classifying the contents of Cresci data set [1] as generated by a bot or a human, where the most accurate prior work achieved an accuracy of 92%.
{"title":"Using BERT to Extract Topic-Independent Sentiment Features for Social Media Bot Detection","authors":"Maryam Heidari, James H. Jones","doi":"10.1109/UEMCON51285.2020.9298158","DOIUrl":"https://doi.org/10.1109/UEMCON51285.2020.9298158","url":null,"abstract":"Millions of online posts about different topics and products are shared on popular social media platforms. One use of this content is to provide crowd-sourced information about a specific topic, event, or product. However, this use raises an important question: what percentage of the information available through these services is trustworthy? In particular, might some of this information be generated by a machine, i.e., a \"bot\" instead of a human? Bots can be, and often are, purposely designed to generate enough volume to skew an apparent trend or position on a topic, yet the consumer of such content cannot easily distinguish a bot post from a human post. This paper introduces a new model that uses Bidirectional Encoder Representations from Transformers (Google Bert) for sentiment classification of tweets to identify topic-independent features for the social media bot detection model. Using a Natural Language Processing approach to derive topic-independent features for the new bot detection model distinguishes this work from previous bot detection models. We achieve 94% accuracy classifying the contents of Cresci data set [1] as generated by a bot or a human, where the most accurate prior work achieved an accuracy of 92%.","PeriodicalId":433609,"journal":{"name":"2020 11th IEEE Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116274076","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：