Search engines, an information retrieval tool are the main source of information for users’ information need now a day. For every query, the search engine explores its repository and/or indexer to find the relevant documents/URLs for that query. Page ranking algorithms rank the Uniform Resource Locator in abstract section (URLs) according to its relevancy with respect to users’ query. It is analyzed that many of the queries fired by users on search engines are duplicate. There is a scope to improve the performance of search engine to reduce its efforts for duplicate queries. In this paper a proxy server is created that keep store the search results of user queries in web log. The proposed proxy server uses this web log to find results faster for duplicate queries fired next time. The proposed scheme has been tested and found prominent. The proposed architecture tested for ten duplicate user queries. it return all relevant web pages for duplicate user query (if query is found in web log at proxy server) from a particular domain instead of entire database. It reduces the perceived latency for duplicate query and also improves the value of precession and accuracy up to 81.8% and 99% respectively for all duplicate user queries.
{"title":"A Novel Architecture for Search Engine using Domain Based Web Log Data","authors":"P. Sharma, Divakar Yadav","doi":"10.34028/iajit/20/1/10","DOIUrl":"https://doi.org/10.34028/iajit/20/1/10","url":null,"abstract":"Search engines, an information retrieval tool are the main source of information for users’ information need now a day. For every query, the search engine explores its repository and/or indexer to find the relevant documents/URLs for that query. Page ranking algorithms rank the Uniform Resource Locator in abstract section (URLs) according to its relevancy with respect to users’ query. It is analyzed that many of the queries fired by users on search engines are duplicate. There is a scope to improve the performance of search engine to reduce its efforts for duplicate queries. In this paper a proxy server is created that keep store the search results of user queries in web log. The proposed proxy server uses this web log to find results faster for duplicate queries fired next time. The proposed scheme has been tested and found prominent. The proposed architecture tested for ten duplicate user queries. it return all relevant web pages for duplicate user query (if query is found in web log at proxy server) from a particular domain instead of entire database. It reduces the perceived latency for duplicate query and also improves the value of precession and accuracy up to 81.8% and 99% respectively for all duplicate user queries.","PeriodicalId":13624,"journal":{"name":"Int. Arab J. Inf. Technol.","volume":"8 1","pages":"92-101"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91047162","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}
The Internet of Things (IoT) is widely known as a revolutionary paradigm that offers communication among different types of devices. The primary goal of this paradigm is to implement efficient and high-quality smart services. It requires a protocol stack that offers different service requirements for inter-communication between different devices. Transmission Control Protocol (TCP) and User Datagram Protocol (UDP) are used as transport layer protocols in IoT to provide the quality of service needed in various IoT devices. IoT encounters many shortcomings of wireless networks, while also posing new challenges due to its uniqueness. When TCP is used in an IoT system, a variety of challenging issues have to be dealt with. This paper provides a comprehensive survey of various issues which arises due to the heterogeneous characteristics of IoT. We identify main issues such as Retransmission Timeout (RTO) algorithm issue, congestion and packet loss issue, header overhead, high latency issue, link layer interaction issue, etc. Moreover, we provide several most probable solutions to the above-mentioned issues in the case of IoT scenarios. RTO algorithm issue has been resolved by using algorithms such as CoCoA, CoCoA+, and CoCoA++. Apart from these, the high latency issue has been solved with the help of a long lived connection and TCP Fast open. Congestion and packet loss issue has been resolved by using several TCP variants such as TCP New Reno, Tahoe, Reno, Vegas, and Westwood.
物联网(IoT)被广泛认为是一种革命性的范例，它提供了不同类型设备之间的通信。此范例的主要目标是实现高效和高质量的智能服务。它需要一个协议栈，为不同设备之间的相互通信提供不同的服务需求。传输控制协议(TCP)和用户数据报协议(UDP)作为物联网中的传输层协议，为各种物联网设备提供所需的服务质量。物联网遇到了无线网络的许多缺点，同时也因其独特性提出了新的挑战。在物联网系统中使用TCP时，必须处理各种具有挑战性的问题。本文全面调查了由于物联网的异构特性而产生的各种问题。我们确定了主要问题，如重传超时(RTO)算法问题，拥塞和数据包丢失问题，报头开销，高延迟问题，链路层交互问题等。此外，在物联网场景下，我们为上述问题提供了几种最可能的解决方案。RTO算法问题已经通过使用CoCoA、CoCoA+和CoCoA+等算法解决。除此之外，在长连接和TCP快速打开的帮助下，高延迟问题已经得到了解决。拥塞和丢包问题已经通过使用几个TCP变体(如TCP New Reno、Tahoe、Reno、Vegas和Westwood)得到解决。
{"title":"Analysis of TCP issues and their possible solutions in the internet of things","authors":"S. Z. Hussain, Sultana Parween","doi":"10.34028/iajit/20/2/7","DOIUrl":"https://doi.org/10.34028/iajit/20/2/7","url":null,"abstract":"The Internet of Things (IoT) is widely known as a revolutionary paradigm that offers communication among different types of devices. The primary goal of this paradigm is to implement efficient and high-quality smart services. It requires a protocol stack that offers different service requirements for inter-communication between different devices. Transmission Control Protocol (TCP) and User Datagram Protocol (UDP) are used as transport layer protocols in IoT to provide the quality of service needed in various IoT devices. IoT encounters many shortcomings of wireless networks, while also posing new challenges due to its uniqueness. When TCP is used in an IoT system, a variety of challenging issues have to be dealt with. This paper provides a comprehensive survey of various issues which arises due to the heterogeneous characteristics of IoT. We identify main issues such as Retransmission Timeout (RTO) algorithm issue, congestion and packet loss issue, header overhead, high latency issue, link layer interaction issue, etc. Moreover, we provide several most probable solutions to the above-mentioned issues in the case of IoT scenarios. RTO algorithm issue has been resolved by using algorithms such as CoCoA, CoCoA+, and CoCoA++. Apart from these, the high latency issue has been solved with the help of a long lived connection and TCP Fast open. Congestion and packet loss issue has been resolved by using several TCP variants such as TCP New Reno, Tahoe, Reno, Vegas, and Westwood.","PeriodicalId":13624,"journal":{"name":"Int. Arab J. Inf. Technol.","volume":"16 1","pages":"206-214"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81109562","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}
Software Defined Wireless Sensor Networks (SDWSN) is emerged to overcome the additional energy consumption in WSN. Even then the sensor nodes in the SDWSN suffer from scarce battery resources. Generally, the Radio Frequency (RF) transmitters are deployed around the base station in the SDWSN to overcome the high energy consumption problem. To enhance harvesting energy and coverage of nodes in the network, a new energy harvesting technique using RF transmitters with Multiple Input and Single Output (MISO) beamforming is proposed. In this method, multiple antenna RF transmitters and single antenna sensor nodes are deployed. The optimization problem subject to Signal to Noise Ratio (SNR) and energy harvesting constraints is formulated for hybrid beamforming design to reduce the transmit power in the network. The optimization problem based on convex Second Order Cone Programming (SOCP) is derived to get the optimal solution for hybrid beamforming design. The beamforming technique is used to steer the beam in the desired direction and null to the other direction improves the energy harvesting. The simulation results show that the proposed technique provides better average harvesting energy, average transmit power, average residual energy and throughput than the existing RF transmitter based energy harvesting methods.
{"title":"A Novel Energy Efficient Harvesting Technique for SDWSN using RF Transmitters with MISO Beamforming","authors":"Subaselvi Sundarraj, Gunaseelan Konganathan","doi":"10.34028/iajit/20/1/13","DOIUrl":"https://doi.org/10.34028/iajit/20/1/13","url":null,"abstract":"Software Defined Wireless Sensor Networks (SDWSN) is emerged to overcome the additional energy consumption in WSN. Even then the sensor nodes in the SDWSN suffer from scarce battery resources. Generally, the Radio Frequency (RF) transmitters are deployed around the base station in the SDWSN to overcome the high energy consumption problem. To enhance harvesting energy and coverage of nodes in the network, a new energy harvesting technique using RF transmitters with Multiple Input and Single Output (MISO) beamforming is proposed. In this method, multiple antenna RF transmitters and single antenna sensor nodes are deployed. The optimization problem subject to Signal to Noise Ratio (SNR) and energy harvesting constraints is formulated for hybrid beamforming design to reduce the transmit power in the network. The optimization problem based on convex Second Order Cone Programming (SOCP) is derived to get the optimal solution for hybrid beamforming design. The beamforming technique is used to steer the beam in the desired direction and null to the other direction improves the energy harvesting. The simulation results show that the proposed technique provides better average harvesting energy, average transmit power, average residual energy and throughput than the existing RF transmitter based energy harvesting methods.","PeriodicalId":13624,"journal":{"name":"Int. Arab J. Inf. Technol.","volume":"167 ","pages":"125-133"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72555248","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}
In recent years, the quantities and types of malwares have grown explosively, which bring many challenges to identify and detect them. In order to improve the identification efficiency of malicious code, a malicious code feature representation method based on feature dimension reduction is proposed. By fusing the Gini impurity increment and the Improved Term Frequency-Inverse Document Frequency algorithm (ITF-IDF), ΔGini-Improving Term frequency inverse document frequency (ΔGini-ITFIDF) method is presented, which can get more valuable assembly instruction features for family detection. ΔGini-ITFIDF standardizes the assembly instructions of the PE disassembly files, then, measures the two indicators of the expected error rate increment and weight of the malicious code assembly instruction features, and obtains more valuable features to identify malicious codes. The experimental results show that the classification accuracy of the ITF-IDF algorithm is significantly improved compared with the ITF-IDF algorithm. At the same time, ΔGini-ITFIDF can effectively improve the classification performance.
{"title":"A method of extracting malware features based on gini impurity increment and improved TF-IDF","authors":"Shimiao Sun, Yashu Liu","doi":"10.34028/iajit/20/3/14","DOIUrl":"https://doi.org/10.34028/iajit/20/3/14","url":null,"abstract":"In recent years, the quantities and types of malwares have grown explosively, which bring many challenges to identify and detect them. In order to improve the identification efficiency of malicious code, a malicious code feature representation method based on feature dimension reduction is proposed. By fusing the Gini impurity increment and the Improved Term Frequency-Inverse Document Frequency algorithm (ITF-IDF), ΔGini-Improving Term frequency inverse document frequency (ΔGini-ITFIDF) method is presented, which can get more valuable assembly instruction features for family detection. ΔGini-ITFIDF standardizes the assembly instructions of the PE disassembly files, then, measures the two indicators of the expected error rate increment and weight of the malicious code assembly instruction features, and obtains more valuable features to identify malicious codes. The experimental results show that the classification accuracy of the ITF-IDF algorithm is significantly improved compared with the ITF-IDF algorithm. At the same time, ΔGini-ITFIDF can effectively improve the classification performance.","PeriodicalId":13624,"journal":{"name":"Int. Arab J. Inf. Technol.","volume":"48 1","pages":"419-427"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75584257","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}
Gas-liquid two-phase flow is widely involved in many scientific and technological fields, such as energy, electricity, nuclear energy, aerospace and environmental protection. In some fields, extracting the accurate position of bubbles in space can not only accurately capture the characteristics of bubbles in two-phase flow, but also plays an important role in the subsequent research like bubble tracking. It has got some progresses to use Convolutional Neural Network (CNNs) to identify bubbles in gas-liquid two-phase flow, while accurate pixel segmentation map in the bubble identification problem is more desirable in many areas. In this paper, VGG16-FCN model and U-Net model are utilized to identify bubbles in two-phase flow images from the perspective of semantic segmentation. LabelMe is used to label the images in the experiment, which can remove the noise in the original image. In addition, bubble pixels with low ratio relative to the background affects the loss function value tinily which cause the irrational evaluation for the recognition in traditional semantic segmentation, thus, Dice loss is used as the loss function for training to improve the recognition effect. The research results show that the two deep learning models have strong feature extraction ability and accurately detect the bubble boundary.
{"title":"Intelligent recognition of gas-liquid two-phase flow based on optical image","authors":"Shujuan Wang, Haofu Guan, Yuqing Wang, Kanghui Zhang, Yuntao Dai, S. Qiao","doi":"10.34028/iajit/20/4/7","DOIUrl":"https://doi.org/10.34028/iajit/20/4/7","url":null,"abstract":"Gas-liquid two-phase flow is widely involved in many scientific and technological fields, such as energy, electricity, nuclear energy, aerospace and environmental protection. In some fields, extracting the accurate position of bubbles in space can not only accurately capture the characteristics of bubbles in two-phase flow, but also plays an important role in the subsequent research like bubble tracking. It has got some progresses to use Convolutional Neural Network (CNNs) to identify bubbles in gas-liquid two-phase flow, while accurate pixel segmentation map in the bubble identification problem is more desirable in many areas. In this paper, VGG16-FCN model and U-Net model are utilized to identify bubbles in two-phase flow images from the perspective of semantic segmentation. LabelMe is used to label the images in the experiment, which can remove the noise in the original image. In addition, bubble pixels with low ratio relative to the background affects the loss function value tinily which cause the irrational evaluation for the recognition in traditional semantic segmentation, thus, Dice loss is used as the loss function for training to improve the recognition effect. The research results show that the two deep learning models have strong feature extraction ability and accurately detect the bubble boundary.","PeriodicalId":13624,"journal":{"name":"Int. Arab J. Inf. Technol.","volume":"263 1","pages":"609-617"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76772650","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}
An Abdominal Aorta Aneurysm (AAA) is an abnormal focal dilation of the aorta. Most un-ruptured AAAs are asymptomatic, which leads to the problem of having abdominal malignancy, kidney damage, heart attack and even death. As it is ominous, it requires an astute scrutinizing approach. The significance of this proposed work is to scrutinize the exact location of the ruptured region and to make astute report of the pathological condition of AAA by computing the Ruptured Potential Index (RPI). To determine these two factors, image processing is performed in the retrieved image of aneurysm. Initially, it undergoes a process to obtain a high-quality image by making use of Adaptive median filter. After retrieving high quality image, segmentation is carried out using Artificial Neural Network-based segmentation. After segmenting the image into samples, 12 features are extracted from the segmented image by Gray Level Co-Occurrence Matrix (GLCM), which assists in extracting the best feature out of it. This optimization is performed by using Particle Swarm Optimization (PSO). Finally, Grey Neural Network (GNN) classifier is applied to analogize the trained and test set data. This classifier helps to achieve the targeted objective with high accuracy.
{"title":"Highly accurate grey neural network classifier for an abdominal aortic aneurysm classification based on image processing approach","authors":"A. Bose, Vasuki Ramesh","doi":"10.34028/iajit/20/2/8","DOIUrl":"https://doi.org/10.34028/iajit/20/2/8","url":null,"abstract":"An Abdominal Aorta Aneurysm (AAA) is an abnormal focal dilation of the aorta. Most un-ruptured AAAs are asymptomatic, which leads to the problem of having abdominal malignancy, kidney damage, heart attack and even death. As it is ominous, it requires an astute scrutinizing approach. The significance of this proposed work is to scrutinize the exact location of the ruptured region and to make astute report of the pathological condition of AAA by computing the Ruptured Potential Index (RPI). To determine these two factors, image processing is performed in the retrieved image of aneurysm. Initially, it undergoes a process to obtain a high-quality image by making use of Adaptive median filter. After retrieving high quality image, segmentation is carried out using Artificial Neural Network-based segmentation. After segmenting the image into samples, 12 features are extracted from the segmented image by Gray Level Co-Occurrence Matrix (GLCM), which assists in extracting the best feature out of it. This optimization is performed by using Particle Swarm Optimization (PSO). Finally, Grey Neural Network (GNN) classifier is applied to analogize the trained and test set data. This classifier helps to achieve the targeted objective with high accuracy.","PeriodicalId":13624,"journal":{"name":"Int. Arab J. Inf. Technol.","volume":"112 1","pages":"215-223"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79666666","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}
One of the basic characteristics of a Gabor filter is that it provides useful information about specific frequencies in a localized region. Such information can be used in locating snippets of code, i.e., localized code, in a program when transformed into an image for finding embedded malicious patterns. Keeping this phenomenon, we propose a novel technique using a sliding Window over Gabor filters for mining the Dalvik Executable (DEX) bytecodes of an Android application (APK) to find malicious patterns. We extract the structural and behavioral functionality and localized information of an APK through Gabor filtered images of the 2D grayscale image of the DEX bytecodes. A Window is slid over these features and a weight is assigned based on its frequency of use. The selected Windows whose weights are greater than a given threshold, are used for training a classifier to detect malware APKs. Our technique does not require any disassembly or execution of the malware program and hence is much safer and more accurate. To further improve feature selection, we apply a greedy optimization algorithm to find the best performing feature subset. The proposed technique, when tested using real malware and benign APKs, obtained a detection rate of 98.9% with 10-fold cross-validation.
{"title":"Mining android bytecodes through the eyes of gabor filters for detecting malware","authors":"Shahid Alam, A. K. Demir","doi":"10.34028/iajit/20/2/4","DOIUrl":"https://doi.org/10.34028/iajit/20/2/4","url":null,"abstract":"One of the basic characteristics of a Gabor filter is that it provides useful information about specific frequencies in a localized region. Such information can be used in locating snippets of code, i.e., localized code, in a program when transformed into an image for finding embedded malicious patterns. Keeping this phenomenon, we propose a novel technique using a sliding Window over Gabor filters for mining the Dalvik Executable (DEX) bytecodes of an Android application (APK) to find malicious patterns. We extract the structural and behavioral functionality and localized information of an APK through Gabor filtered images of the 2D grayscale image of the DEX bytecodes. A Window is slid over these features and a weight is assigned based on its frequency of use. The selected Windows whose weights are greater than a given threshold, are used for training a classifier to detect malware APKs. Our technique does not require any disassembly or execution of the malware program and hence is much safer and more accurate. To further improve feature selection, we apply a greedy optimization algorithm to find the best performing feature subset. The proposed technique, when tested using real malware and benign APKs, obtained a detection rate of 98.9% with 10-fold cross-validation.","PeriodicalId":13624,"journal":{"name":"Int. Arab J. Inf. Technol.","volume":"33 1","pages":"180-189"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80422811","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}
In Wireless Sensor Networks (WSNs), clustering is an effective method to distribute the load equally among all the nodes as compared to flat network architecture. Due to the dynamic nature of the network, the clustering process can be viewed as a dynamic optimization problem and the conventional computational intelligence techniques are not enough to solve these problems. The Dynamic Genetic Algorithm (DGA) addresses these problems with the help of searching optimal solutions in new environments. Therefore the dynamic load-balanced clustering process is modeled using the basic components of standard genetic algorithm and then the model is enhanced is using immigrants and memory-based schemes to elect suitable cluster heads. The metrics nodes’ residual energy level, node centrality, and mobility speed of the nodes are considered to elect the load-balanced cluster heads and the optimal number of cluster members are assigned to each cluster head using the proposed DGA schemes such as Random Immigrants Genetic Approach (RIGA), Memory Immigrants Genetic Approach (MIGA), and Memory and Random Immigrants Genetic Approach (MRIGA). The simulation results show that the proposed DGA scheme MRIGA outperforms well as compared with RIGA and MIGA in terms of various performance metrics such as the number of nodes alive, residual energy level, packet delivery ratio, end-to-end delay, and overhead for the formation of clusters.
{"title":"Genetic algorithm with random and memory immigrant strategies for solving dynamic load balanced clustering problem in wireless sensor networks","authors":"Mohaideen Pitchai","doi":"10.34028/iajit/20/4/3","DOIUrl":"https://doi.org/10.34028/iajit/20/4/3","url":null,"abstract":"In Wireless Sensor Networks (WSNs), clustering is an effective method to distribute the load equally among all the nodes as compared to flat network architecture. Due to the dynamic nature of the network, the clustering process can be viewed as a dynamic optimization problem and the conventional computational intelligence techniques are not enough to solve these problems. The Dynamic Genetic Algorithm (DGA) addresses these problems with the help of searching optimal solutions in new environments. Therefore the dynamic load-balanced clustering process is modeled using the basic components of standard genetic algorithm and then the model is enhanced is using immigrants and memory-based schemes to elect suitable cluster heads. The metrics nodes’ residual energy level, node centrality, and mobility speed of the nodes are considered to elect the load-balanced cluster heads and the optimal number of cluster members are assigned to each cluster head using the proposed DGA schemes such as Random Immigrants Genetic Approach (RIGA), Memory Immigrants Genetic Approach (MIGA), and Memory and Random Immigrants Genetic Approach (MRIGA). The simulation results show that the proposed DGA scheme MRIGA outperforms well as compared with RIGA and MIGA in terms of various performance metrics such as the number of nodes alive, residual energy level, packet delivery ratio, end-to-end delay, and overhead for the formation of clusters.","PeriodicalId":13624,"journal":{"name":"Int. Arab J. Inf. Technol.","volume":"11 19 1","pages":"575-583"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87226218","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}
Retina disorders are among the common types of eye disease that occur due to several reasons such as aging, diabetes and premature born. Besides, Optical Coherence Tomography (OCT) is a medical imaging method that serves as a vehicle for capturing volumetric scans of the human eye retina for diagnoses purposes. This research compared two pretraining approaches including Self-Supervised Learning (SSL) and Transfer Learning (TL) to train ResNet34 neural architecture aiming at building computer aided diagnoses tool for retina disorders recognition. In addition, the research methodology employs convolutional auto-encoder model as a generative SSL pretraining method. The research efforts are implemented on a dataset that contains 109,309 retina OCT images with three medical conditions including Choroidal Neovascularization (CNV), Diabetic Macular Edema (DME), DRUSEN as well as NORMAL condition. The research outcomes showed better performance in terms of overall accuracy, sensitivity and specificity, namely, 95.2%, 95.2% and 98.4% respectively for SSL ResNet34 in comparison to scores of 90.7%, 90.7% and 96.9% respectively for TL ResNet34. In addition, SSL pretraining approach showed significant reduction in the number of epochs required for training in comparison to both TL pretraining as well as the previous research performed on the same dataset with comparable performance.
{"title":"Retina disorders classification via OCT scan: a comparative study between self-supervised learning and transfer learning","authors":"Saeed Shurrab, Yazan Shannak, R. Duwairi","doi":"10.34028/iajit/20/3/8","DOIUrl":"https://doi.org/10.34028/iajit/20/3/8","url":null,"abstract":"Retina disorders are among the common types of eye disease that occur due to several reasons such as aging, diabetes and premature born. Besides, Optical Coherence Tomography (OCT) is a medical imaging method that serves as a vehicle for capturing volumetric scans of the human eye retina for diagnoses purposes. This research compared two pretraining approaches including Self-Supervised Learning (SSL) and Transfer Learning (TL) to train ResNet34 neural architecture aiming at building computer aided diagnoses tool for retina disorders recognition. In addition, the research methodology employs convolutional auto-encoder model as a generative SSL pretraining method. The research efforts are implemented on a dataset that contains 109,309 retina OCT images with three medical conditions including Choroidal Neovascularization (CNV), Diabetic Macular Edema (DME), DRUSEN as well as NORMAL condition. The research outcomes showed better performance in terms of overall accuracy, sensitivity and specificity, namely, 95.2%, 95.2% and 98.4% respectively for SSL ResNet34 in comparison to scores of 90.7%, 90.7% and 96.9% respectively for TL ResNet34. In addition, SSL pretraining approach showed significant reduction in the number of epochs required for training in comparison to both TL pretraining as well as the previous research performed on the same dataset with comparable performance.","PeriodicalId":13624,"journal":{"name":"Int. Arab J. Inf. Technol.","volume":"28 1","pages":"357-367"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84319388","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}
The ability to post short text and media messages on Social media platforms like Twitter, Facebook, etc., plays a huge role in the exchange of information following a mass emergency event like hurricane, earthquake, tsunami etc. Disaster victims, families, and other relief operation teams utilize social media to help and support one another. Despite the benefits offered by these communication media, the disaster topic related posts (posts that indicate conversations about the disaster event in the aftermath of the disaster) gets lost in the deluge of posts since there would be a surge in the amount of data that gets exchanged following a mass emergency event. This hampers the emergency relief effort, which in turn affects the delivery of useful information to the disaster victims. Research in emergency coordination via social media has received growing interest in recent years, mainly focusing on developing machine learning-based models that can separate disaster-related topic posts from non-disaster related topic posts. Of these, supervised machine learning approaches performed well when the machine learning model trained using source disaster dataset and target disaster dataset are similar. However, in the real world, it may not be feasible as different disasters have different characteristics. So, models developed using supervised machine learning approaches do not perform well in unseen disaster datasets. Therefore, domain adaptation approaches, which address the above limitation by learning classifiers from unlabeled target data in addition to source labelled data, represent a promising direction for social media crisis data classification tasks. The existing domain adaptation techniques for the classification of disaster tweets are experimented with using single disaster event dataset pairs; then, self-training is performed on the source target dataset pairs by considering the highly confident instances in subsequent iterations of training. This could be improved with better feature engineering. Thus, this research proposes a Genetic Algorithm based Domain Adaptation Framework (GADA) for the classification of disaster tweets. The proposed GADA combines the power of 1) Hybrid Feature Selection component using the Genetic Algorithm and Chi-Square Feature Evaluator for feature selection and 2) the Classifier component using Random Forest to classify disaster-related posts from noise on Twitter. The proposed framework addresses the challenge of the lack of labeled data in the target disaster event by proposing a Genetic Algorithm based approach. Experimental results on Twitter datasets corresponding to four disaster domain pair shows that the proposed framework improves the overall performance of the previous supervised approaches and significantly reduces the training time over the previous domain adaptation techniques that do not use the Genetic Algorithm (GA) for feature selection.
{"title":"A Genetic Algorithm based Domain Adaptation Framework for Classification of Disaster Topic Text Tweets","authors":"Lokabhiram Dwarakanath, A. Kamsin, Liyana Shuib","doi":"10.34028/iajit/20/1/7","DOIUrl":"https://doi.org/10.34028/iajit/20/1/7","url":null,"abstract":"The ability to post short text and media messages on Social media platforms like Twitter, Facebook, etc., plays a huge role in the exchange of information following a mass emergency event like hurricane, earthquake, tsunami etc. Disaster victims, families, and other relief operation teams utilize social media to help and support one another. Despite the benefits offered by these communication media, the disaster topic related posts (posts that indicate conversations about the disaster event in the aftermath of the disaster) gets lost in the deluge of posts since there would be a surge in the amount of data that gets exchanged following a mass emergency event. This hampers the emergency relief effort, which in turn affects the delivery of useful information to the disaster victims. Research in emergency coordination via social media has received growing interest in recent years, mainly focusing on developing machine learning-based models that can separate disaster-related topic posts from non-disaster related topic posts. Of these, supervised machine learning approaches performed well when the machine learning model trained using source disaster dataset and target disaster dataset are similar. However, in the real world, it may not be feasible as different disasters have different characteristics. So, models developed using supervised machine learning approaches do not perform well in unseen disaster datasets. Therefore, domain adaptation approaches, which address the above limitation by learning classifiers from unlabeled target data in addition to source labelled data, represent a promising direction for social media crisis data classification tasks. The existing domain adaptation techniques for the classification of disaster tweets are experimented with using single disaster event dataset pairs; then, self-training is performed on the source target dataset pairs by considering the highly confident instances in subsequent iterations of training. This could be improved with better feature engineering. Thus, this research proposes a Genetic Algorithm based Domain Adaptation Framework (GADA) for the classification of disaster tweets. The proposed GADA combines the power of 1) Hybrid Feature Selection component using the Genetic Algorithm and Chi-Square Feature Evaluator for feature selection and 2) the Classifier component using Random Forest to classify disaster-related posts from noise on Twitter. The proposed framework addresses the challenge of the lack of labeled data in the target disaster event by proposing a Genetic Algorithm based approach. Experimental results on Twitter datasets corresponding to four disaster domain pair shows that the proposed framework improves the overall performance of the previous supervised approaches and significantly reduces the training time over the previous domain adaptation techniques that do not use the Genetic Algorithm (GA) for feature selection.","PeriodicalId":13624,"journal":{"name":"Int. Arab J. Inf. Technol.","volume":"53 1","pages":"57-65"},"PeriodicalIF":0.0,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91358427","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：