Pub Date : 2022-12-02DOI: 10.1109/SKIMA57145.2022.10029617
S. Hul
The Royal Government of Cambodia's aspiration to bring the country into the upper-middle income and high-income groups by 2030 and 2050, respectively, also means that Cambodia will need to persistently build a knowledge-based society and generate more skilled labor to meet the demand from industries and market which are also favored by technological changes. This roadmap's vision is to build the next-generation technology-enhanced learning ecosystem focusing on improving innovation and entrepreneurship skills. Its main objective is to infuse technology rapidly in teaching and learning at home and school to support students acquiring skills and knowledge that they will need for a successful education, life, and career in the modern workplace and society. In this regard, this EduTech roadmap objective is also in line with that of the Cambodia's Science, Technology & Innovation Roadmap 2030. In particular, the EduTech roadmap aims to promote overall learning outcome, digital literacy, entrepreneurship skills, and technological readiness of the Cambodian population, starting from the very young age when children begin their formal general education. From the social, technological, economic, environmental & political (STEEP) approach and SWOT analyses which were done by desk review method and the data that the research team collected through in-depth interviews and roundtable panel discussions and the consensus among EduTech committee members and participants during the consultation and validation workshops, four pillars/areas of products/services have been identified, namely the Essentials, Management System, Courseware, and Capacity Building. Appraisal of importance of each product or service is also based on a few factors, including economic impacts, current and future market potentials, and overall national development agenda.
{"title":"Cambodia: Innovation-Driven EduTech Roadmap 2030","authors":"S. Hul","doi":"10.1109/SKIMA57145.2022.10029617","DOIUrl":"https://doi.org/10.1109/SKIMA57145.2022.10029617","url":null,"abstract":"The Royal Government of Cambodia's aspiration to bring the country into the upper-middle income and high-income groups by 2030 and 2050, respectively, also means that Cambodia will need to persistently build a knowledge-based society and generate more skilled labor to meet the demand from industries and market which are also favored by technological changes. This roadmap's vision is to build the next-generation technology-enhanced learning ecosystem focusing on improving innovation and entrepreneurship skills. Its main objective is to infuse technology rapidly in teaching and learning at home and school to support students acquiring skills and knowledge that they will need for a successful education, life, and career in the modern workplace and society. In this regard, this EduTech roadmap objective is also in line with that of the Cambodia's Science, Technology & Innovation Roadmap 2030. In particular, the EduTech roadmap aims to promote overall learning outcome, digital literacy, entrepreneurship skills, and technological readiness of the Cambodian population, starting from the very young age when children begin their formal general education. From the social, technological, economic, environmental & political (STEEP) approach and SWOT analyses which were done by desk review method and the data that the research team collected through in-depth interviews and roundtable panel discussions and the consensus among EduTech committee members and participants during the consultation and validation workshops, four pillars/areas of products/services have been identified, namely the Essentials, Management System, Courseware, and Capacity Building. Appraisal of importance of each product or service is also based on a few factors, including economic impacts, current and future market potentials, and overall national development agenda.","PeriodicalId":277436,"journal":{"name":"2022 14th International Conference on Software, Knowledge, Information Management and Applications (SKIMA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128428503","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 : 2022-12-02DOI: 10.1109/SKIMA57145.2022.10029514
Bishal Rimal, N. Paudel, Aayush Bhattarai
Optimal Phasor measurement unit (PMU) placement is the strategic placement of a minimum number of PMUs to ensure power system observability. Optimal PMU placement (OPP) is essential to overcome the economic burden during the deployment of PMU on every bus. A strategy integrated with the modified Simulated Annealing (SA) algorithm is proposed in this paper with the consideration of the effect of the radial buses during the OPP problem. To recognize the superior placement set from diverse solutions given by the meta-heuristic algorithm, a placement set with the least number of PMUs and a higher system observability redundancy index (SORI) is used. In addition to modified SA, the strategy of ensuring or restricting the PMU placement on the radial bus has been implemented in MATLAB. In addition to the normal case, cases for single PMU outage and zero injection bus (ZIB) have been implemented in IEEE 14-bus, 30-bus, 57-bus, and 118-bus systems. The algorithm has been also used on the existing 90-bus Integrated Nepal Power System. The result of the proposed strategy on modified SA has improved the result in terms of the optimal number of PMUs, the probability of finding an optimal number of PMUs, and SORI. For normal cases, the number of PMUs required for complete power system observability is found to be in the range of 28-33% of the total number of buses. Similarly, with the consideration of single PMU outages and ZIB, the number of PMUs required is found to be in the range of 57-70% and 21-26% respectively.
{"title":"Optimal Placement of Phasor Measurement Units Ensuring Power System Observability","authors":"Bishal Rimal, N. Paudel, Aayush Bhattarai","doi":"10.1109/SKIMA57145.2022.10029514","DOIUrl":"https://doi.org/10.1109/SKIMA57145.2022.10029514","url":null,"abstract":"Optimal Phasor measurement unit (PMU) placement is the strategic placement of a minimum number of PMUs to ensure power system observability. Optimal PMU placement (OPP) is essential to overcome the economic burden during the deployment of PMU on every bus. A strategy integrated with the modified Simulated Annealing (SA) algorithm is proposed in this paper with the consideration of the effect of the radial buses during the OPP problem. To recognize the superior placement set from diverse solutions given by the meta-heuristic algorithm, a placement set with the least number of PMUs and a higher system observability redundancy index (SORI) is used. In addition to modified SA, the strategy of ensuring or restricting the PMU placement on the radial bus has been implemented in MATLAB. In addition to the normal case, cases for single PMU outage and zero injection bus (ZIB) have been implemented in IEEE 14-bus, 30-bus, 57-bus, and 118-bus systems. The algorithm has been also used on the existing 90-bus Integrated Nepal Power System. The result of the proposed strategy on modified SA has improved the result in terms of the optimal number of PMUs, the probability of finding an optimal number of PMUs, and SORI. For normal cases, the number of PMUs required for complete power system observability is found to be in the range of 28-33% of the total number of buses. Similarly, with the consideration of single PMU outages and ZIB, the number of PMUs required is found to be in the range of 57-70% and 21-26% respectively.","PeriodicalId":277436,"journal":{"name":"2022 14th International Conference on Software, Knowledge, Information Management and Applications (SKIMA)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132259111","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 : 2022-12-02DOI: 10.1109/SKIMA57145.2022.10029459
M. Joseph, Sri Devi Ravana
The Information Retrieval System Evaluation have carried out through Cranfield-paradigm in which the test collections provide the foundation of the evaluation process. The test collections consist of document corpus, topics, and a set of relevance judgements. The relevant judgements are the documents which retrieved from the test collections based on the topics. The precision of the evaluation process is based on the number of relevant documents in the relevant judgement list called qrels. This paper presents a study on how methodologies like pooling and document similarity helps to generate more relevant documents into the relevance judgments set in order to increase the accuracy of the evaluation process. The initial results have shown that combination of pooling with document similarity performs better compared to base clustering or classification.
{"title":"Generation of High-Quality Relevant Judgments through Document Similarity and Document Pooling for the Evaluation of Information Retrieval Systems","authors":"M. Joseph, Sri Devi Ravana","doi":"10.1109/SKIMA57145.2022.10029459","DOIUrl":"https://doi.org/10.1109/SKIMA57145.2022.10029459","url":null,"abstract":"The Information Retrieval System Evaluation have carried out through Cranfield-paradigm in which the test collections provide the foundation of the evaluation process. The test collections consist of document corpus, topics, and a set of relevance judgements. The relevant judgements are the documents which retrieved from the test collections based on the topics. The precision of the evaluation process is based on the number of relevant documents in the relevant judgement list called qrels. This paper presents a study on how methodologies like pooling and document similarity helps to generate more relevant documents into the relevance judgments set in order to increase the accuracy of the evaluation process. The initial results have shown that combination of pooling with document similarity performs better compared to base clustering or classification.","PeriodicalId":277436,"journal":{"name":"2022 14th International Conference on Software, Knowledge, Information Management and Applications (SKIMA)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114067517","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 : 2022-12-02DOI: 10.1109/SKIMA57145.2022.10029673
Hao Xian Chung, Nazia Hameed, Jérémie Clos, M. Hasan
American sign language (ASL) is a natural language to minimize communication barrier for the people suffering from hearing and speech impairment. However, sign languages are not a common form of communication in society, and it is lacking a sign language translator to ease the communication. Hence, this research proposes an improved visual-based communication framework to translate ASL alphabets in real-time. The proposed framework consists of different steps i.e., image pre-processing, hand segmentation with U-Net, data-augmentation techniques, and classification of American sign language. A n ensemble classification model i s proposed where V GG19, R esNet-50 and MobileNet are used as base models, and 2 fully connected layers are used as meta model. The proposed approach achieved 99.86% accuracy and performed better when compared with existing literature.
{"title":"A Framework of Ensemble CNN Models for Real-Time Sign Language Translation","authors":"Hao Xian Chung, Nazia Hameed, Jérémie Clos, M. Hasan","doi":"10.1109/SKIMA57145.2022.10029673","DOIUrl":"https://doi.org/10.1109/SKIMA57145.2022.10029673","url":null,"abstract":"American sign language (ASL) is a natural language to minimize communication barrier for the people suffering from hearing and speech impairment. However, sign languages are not a common form of communication in society, and it is lacking a sign language translator to ease the communication. Hence, this research proposes an improved visual-based communication framework to translate ASL alphabets in real-time. The proposed framework consists of different steps i.e., image pre-processing, hand segmentation with U-Net, data-augmentation techniques, and classification of American sign language. A n ensemble classification model i s proposed where V GG19, R esNet-50 and MobileNet are used as base models, and 2 fully connected layers are used as meta model. The proposed approach achieved 99.86% accuracy and performed better when compared with existing literature.","PeriodicalId":277436,"journal":{"name":"2022 14th International Conference on Software, Knowledge, Information Management and Applications (SKIMA)","volume":"106 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114134602","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 : 2022-12-02DOI: 10.1109/SKIMA57145.2022.10029566
Vatana Chea, Phal Chea
While digital transformation around the globe has significantly gained momentum and speed, embracing technologies in our daily life is now beyond question. In this regard, we need to ask whether or not individuals are ready to accept and use technologies when provided and what really affects one's technology readiness. The main objective of this study is to understand the role of parents and family background in influencing an undergraduate student's technological readiness. We employ the Ordinary Least Square Regression method on nationally representative data known as the Cambodia Post-Secondary Student Survey 2020 with a sample size of 1,338 individuals. Results support the Technology of Skills Formation concept that parents, especially the mothers, play a crucial role socially and financially in influencing individuals' technology readiness which is measured by the Technology Readiness Index, a composite measurement hardly affected by any short-lived disruption. Moreover, parental education is negatively associated with children's technological readiness holding the wealth effect constant.
{"title":"Family Background as the Determinant of University Student's Technological Readiness: Evidence from Cambodia","authors":"Vatana Chea, Phal Chea","doi":"10.1109/SKIMA57145.2022.10029566","DOIUrl":"https://doi.org/10.1109/SKIMA57145.2022.10029566","url":null,"abstract":"While digital transformation around the globe has significantly gained momentum and speed, embracing technologies in our daily life is now beyond question. In this regard, we need to ask whether or not individuals are ready to accept and use technologies when provided and what really affects one's technology readiness. The main objective of this study is to understand the role of parents and family background in influencing an undergraduate student's technological readiness. We employ the Ordinary Least Square Regression method on nationally representative data known as the Cambodia Post-Secondary Student Survey 2020 with a sample size of 1,338 individuals. Results support the Technology of Skills Formation concept that parents, especially the mothers, play a crucial role socially and financially in influencing individuals' technology readiness which is measured by the Technology Readiness Index, a composite measurement hardly affected by any short-lived disruption. Moreover, parental education is negatively associated with children's technological readiness holding the wealth effect constant.","PeriodicalId":277436,"journal":{"name":"2022 14th International Conference on Software, Knowledge, Information Management and Applications (SKIMA)","volume":"312 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115833025","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 : 2022-12-02DOI: 10.1109/SKIMA57145.2022.10029504
Husnain Rafiq, N. Aslam, B. Issac, R. H. Randhawa
Due to the widespread usage of Android-based smartphones in the current era, Android malware has become a significant concern. From the perspective of t he a dvances in machine learning-based approaches in the previous decade, the research community has shown a dominant interest in applying these to counter Android malware. However, these ML-based classifiers are vulnerable to attacks. An attacker can deliberately fabricate the input application to force the classification algorithm to produce the desired output (evasion attack). In this study, first, w e propose HybridDroid, a n M L-based Android malware classifier trained o n hybrid features a nd optimized using the tree-based pipeline optimization technique (TPOT). Our experiments show that HybriDroid achieves a remarkable detection accuracy of up to 99.2% on a balanced excerpt of 36,000 malware and benign Android apps. Secondly, we explore the effectiveness of the proposed model in adversarial environments. We apply mimicry attacks, feature removal attacks and feature removal with injection attacks on HybriDroid. Our experiments reveal that ML-based malware classifiers are highly vulnerable to adversarial evasion attacks. Finally, we propose future directions to harden the security of ML-based Android malware classifiers in adversarial settings.
{"title":"On Impact of Adversarial Evasion Attacks on ML-based Android Malware Classifier Trained on Hybrid Features","authors":"Husnain Rafiq, N. Aslam, B. Issac, R. H. Randhawa","doi":"10.1109/SKIMA57145.2022.10029504","DOIUrl":"https://doi.org/10.1109/SKIMA57145.2022.10029504","url":null,"abstract":"Due to the widespread usage of Android-based smartphones in the current era, Android malware has become a significant concern. From the perspective of t he a dvances in machine learning-based approaches in the previous decade, the research community has shown a dominant interest in applying these to counter Android malware. However, these ML-based classifiers are vulnerable to attacks. An attacker can deliberately fabricate the input application to force the classification algorithm to produce the desired output (evasion attack). In this study, first, w e propose HybridDroid, a n M L-based Android malware classifier trained o n hybrid features a nd optimized using the tree-based pipeline optimization technique (TPOT). Our experiments show that HybriDroid achieves a remarkable detection accuracy of up to 99.2% on a balanced excerpt of 36,000 malware and benign Android apps. Secondly, we explore the effectiveness of the proposed model in adversarial environments. We apply mimicry attacks, feature removal attacks and feature removal with injection attacks on HybriDroid. Our experiments reveal that ML-based malware classifiers are highly vulnerable to adversarial evasion attacks. Finally, we propose future directions to harden the security of ML-based Android malware classifiers in adversarial settings.","PeriodicalId":277436,"journal":{"name":"2022 14th International Conference on Software, Knowledge, Information Management and Applications (SKIMA)","volume":"242 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132679557","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 : 2022-12-02DOI: 10.1109/SKIMA57145.2022.10029659
M. Hassan, M. A. Hossain
Due to the remarkable advancement of technology, smaller form factors and affordability, immersive technologies such as Virtual Reality (VR) are getting considerable attraction in the research community. The use of VR in various fields, i.e., health care education, gaming, social media etc. is now well established. Among these, perhaps the most dynamic area of VR application is the ‘Skill Training’ sector. VR has been in use for training people for critical tasks (both cognitive and procedural) as it requires less training time and the training cost can also be minimized significantly. Moreover, prior research shows that VR can improve memory retention and reduce human error, which subsequently shows the effectiveness of VR training. In this paper, we have proposed a novel VR training simulation regarding children's formula milk preparation. Though ‘formula milk preparation’ appears to be a trivial task, it indeed possesses substantial dangers in perspective to the child's health and well-being. Because of the trivial nature of the task, there is always a possibility of human errors while performing the baby feed preparation. We argue, through a VR-based training platform, the human error part can be minimized since the trainee will perform the actions procedurally, resulting in greater retention of the suggested process. In addition to this, we have also proposed a novel AI-enabled backend training performance assessor that can assess the performance of trainees while they are trained in VR.
{"title":"A VR based children formula feed preparation training simulator with AI-enabled automated assessment features","authors":"M. Hassan, M. A. Hossain","doi":"10.1109/SKIMA57145.2022.10029659","DOIUrl":"https://doi.org/10.1109/SKIMA57145.2022.10029659","url":null,"abstract":"Due to the remarkable advancement of technology, smaller form factors and affordability, immersive technologies such as Virtual Reality (VR) are getting considerable attraction in the research community. The use of VR in various fields, i.e., health care education, gaming, social media etc. is now well established. Among these, perhaps the most dynamic area of VR application is the ‘Skill Training’ sector. VR has been in use for training people for critical tasks (both cognitive and procedural) as it requires less training time and the training cost can also be minimized significantly. Moreover, prior research shows that VR can improve memory retention and reduce human error, which subsequently shows the effectiveness of VR training. In this paper, we have proposed a novel VR training simulation regarding children's formula milk preparation. Though ‘formula milk preparation’ appears to be a trivial task, it indeed possesses substantial dangers in perspective to the child's health and well-being. Because of the trivial nature of the task, there is always a possibility of human errors while performing the baby feed preparation. We argue, through a VR-based training platform, the human error part can be minimized since the trainee will perform the actions procedurally, resulting in greater retention of the suggested process. In addition to this, we have also proposed a novel AI-enabled backend training performance assessor that can assess the performance of trainees while they are trained in VR.","PeriodicalId":277436,"journal":{"name":"2022 14th International Conference on Software, Knowledge, Information Management and Applications (SKIMA)","volume":"139 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114449749","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 : 2022-12-02DOI: 10.1109/SKIMA57145.2022.10029640
F. B. Setiawan, S. Riyadi, L. Pratomo, A. Wibisono
The microgrid system is a self-contained electrical system that is commonly installed in certain locations such as campuses, business centers, hospitals, residential areas etc. In the microgrid, there are generally distributed energy sources including solar panels, wind turbines etc and also equipped with energy storage elements (generally batteries). It is operated for reliable and efficient operation. The microgrid system for Soegijapranata Catholic University consists of 5.4 kWp photovoltaic arrays and 14,4 kWh batteries which are connected to utility, they are controlled via SCADA system. The microgrid system can operated under grid-connected or inslanding mode with different conditions. This paper describes such a microgrid for professional and higher education training, includes its operation modes.
{"title":"A 5.4 kWp Microgrid Laboratory Development for Higher Education and Industrial Workshop","authors":"F. B. Setiawan, S. Riyadi, L. Pratomo, A. Wibisono","doi":"10.1109/SKIMA57145.2022.10029640","DOIUrl":"https://doi.org/10.1109/SKIMA57145.2022.10029640","url":null,"abstract":"The microgrid system is a self-contained electrical system that is commonly installed in certain locations such as campuses, business centers, hospitals, residential areas etc. In the microgrid, there are generally distributed energy sources including solar panels, wind turbines etc and also equipped with energy storage elements (generally batteries). It is operated for reliable and efficient operation. The microgrid system for Soegijapranata Catholic University consists of 5.4 kWp photovoltaic arrays and 14,4 kWh batteries which are connected to utility, they are controlled via SCADA system. The microgrid system can operated under grid-connected or inslanding mode with different conditions. This paper describes such a microgrid for professional and higher education training, includes its operation modes.","PeriodicalId":277436,"journal":{"name":"2022 14th International Conference on Software, Knowledge, Information Management and Applications (SKIMA)","volume":"517 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116230119","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 : 2022-12-02DOI: 10.1109/SKIMA57145.2022.10029527
M. Wang, W. Leelapatra
In the process of planting, weeds will inevitably grow in the farmland, and compete with crops for water, light and space, which obviously affect our normal agriculture. If weeds are not effectively controlled, crops output will be seriously compromised. On the other hand, it also increases the number of pests. Nowadays, the main weeding methods rely on labor and chemical herbicide. However, manual weeding is inefficient, costly, time consuming and cannot remove weeds effectively. The purpose of this paper is to propose a weeding robot. The focus of the research is on how to use dual cameras to accurately detect weeds. The convolutional neural networks (CNNs), deep learning, dual cameras machine vision and mechanical design will be discussed in this paper. The results show that dual cameras robot based on a new lightweight platform can achieve a high accuracy compared to single camera method, while a feasible rail system was proposed for weeding robots. In vegetable detection, this method achieves 98.12% precision, 83.47% recall and 89.91% mAP that is 4.06% higher than a single top view camera. GF-YOLO, a lightweight platform we proposed also outperform other state-of-the-art algorithms in embedded system.
{"title":"Weeding Robot Based on Lightweight Platform and Dual Cameras","authors":"M. Wang, W. Leelapatra","doi":"10.1109/SKIMA57145.2022.10029527","DOIUrl":"https://doi.org/10.1109/SKIMA57145.2022.10029527","url":null,"abstract":"In the process of planting, weeds will inevitably grow in the farmland, and compete with crops for water, light and space, which obviously affect our normal agriculture. If weeds are not effectively controlled, crops output will be seriously compromised. On the other hand, it also increases the number of pests. Nowadays, the main weeding methods rely on labor and chemical herbicide. However, manual weeding is inefficient, costly, time consuming and cannot remove weeds effectively. The purpose of this paper is to propose a weeding robot. The focus of the research is on how to use dual cameras to accurately detect weeds. The convolutional neural networks (CNNs), deep learning, dual cameras machine vision and mechanical design will be discussed in this paper. The results show that dual cameras robot based on a new lightweight platform can achieve a high accuracy compared to single camera method, while a feasible rail system was proposed for weeding robots. In vegetable detection, this method achieves 98.12% precision, 83.47% recall and 89.91% mAP that is 4.06% higher than a single top view camera. GF-YOLO, a lightweight platform we proposed also outperform other state-of-the-art algorithms in embedded system.","PeriodicalId":277436,"journal":{"name":"2022 14th International Conference on Software, Knowledge, Information Management and Applications (SKIMA)","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123980331","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 : 2022-12-02DOI: 10.1109/SKIMA57145.2022.10029452
N. Dina, Sri Devi Ravana, N. Idris
Text sentiment classification aims to extract useful information from unstructured text data and classify its sentiment into positive and negative categories. Irrelevant features and high-dimensional feature space from text data are common issues in sentiment classification because they degrade the classification performance. To address these issues, this study applies hybrid feature selection using Term Frequency-Inverse Document Frequency (TF-IDF) and Support Vector Machine-Recursive Feature Elimination (SVM-RFE) to three text datasets: IMDB, Yelp, and Amazon. The TF-IDF is employed to select sentiment features, which are further refined by SVM-RFE. Finally, SVM is applied to determine whether the sentiment is positive or negative. This study outperforms the existing techniques in two datasets: 88% accuracy in the IMDB dataset and 84.5% in the Yelp dataset. Meanwhile, the accuracy in the Amazon dataset is lower than the existing studies, at 81.5%. These results indicate inconsistency of the technique, and it opens the opportunity for further research on the other hybrid feature selection techniques for sentiment classification to improve the accuracy in all datasets. Also, the results show that the technique improved classification performance and reduced feature space by 63%.
{"title":"An Experimental Study on Hybrid Feature Selection Techniques for Sentiment Classification","authors":"N. Dina, Sri Devi Ravana, N. Idris","doi":"10.1109/SKIMA57145.2022.10029452","DOIUrl":"https://doi.org/10.1109/SKIMA57145.2022.10029452","url":null,"abstract":"Text sentiment classification aims to extract useful information from unstructured text data and classify its sentiment into positive and negative categories. Irrelevant features and high-dimensional feature space from text data are common issues in sentiment classification because they degrade the classification performance. To address these issues, this study applies hybrid feature selection using Term Frequency-Inverse Document Frequency (TF-IDF) and Support Vector Machine-Recursive Feature Elimination (SVM-RFE) to three text datasets: IMDB, Yelp, and Amazon. The TF-IDF is employed to select sentiment features, which are further refined by SVM-RFE. Finally, SVM is applied to determine whether the sentiment is positive or negative. This study outperforms the existing techniques in two datasets: 88% accuracy in the IMDB dataset and 84.5% in the Yelp dataset. Meanwhile, the accuracy in the Amazon dataset is lower than the existing studies, at 81.5%. These results indicate inconsistency of the technique, and it opens the opportunity for further research on the other hybrid feature selection techniques for sentiment classification to improve the accuracy in all datasets. Also, the results show that the technique improved classification performance and reduced feature space by 63%.","PeriodicalId":277436,"journal":{"name":"2022 14th International Conference on Software, Knowledge, Information Management and Applications (SKIMA)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130534107","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：