Pub Date : 2024-06-01DOI: 10.11591/ijai.v13.i2.pp2441-2450
Imam Bil Qisthi, Hartono Siswono
Nutmeg seeds can produce a lot of oil if they have optimal maturity, in other words, they have little moisture content. Based on observations I made at one of the refineries in Sukabumi, farmers do not pay attention to the maturity level of nutmeg seeds after drying which can cause a decrease in the quality of nutmeg seeds and the quality of the oil produced. This study aims to make it easier for nutmeg farmers to classify the maturity of nutmeg seeds. This study used the Convolutional Neural Network (CNN) method to help with classification problems and several image processing methods. This program will be run through an android application. The results of CNN model training accuracy are 97.92%. Thus, it can be concluded that the design and testing of a model to classify the maturity level of nutmeg seeds using artificial intelligence and the implementation of the model into an android application has been successfully carried out.
{"title":"Classification of nutmeg ripeness using artificial intelligence","authors":"Imam Bil Qisthi, Hartono Siswono","doi":"10.11591/ijai.v13.i2.pp2441-2450","DOIUrl":"https://doi.org/10.11591/ijai.v13.i2.pp2441-2450","url":null,"abstract":"Nutmeg seeds can produce a lot of oil if they have optimal maturity, in other words, they have little moisture content. Based on observations I made at one of the refineries in Sukabumi, farmers do not pay attention to the maturity level of nutmeg seeds after drying which can cause a decrease in the quality of nutmeg seeds and the quality of the oil produced. This study aims to make it easier for nutmeg farmers to classify the maturity of nutmeg seeds. This study used the Convolutional Neural Network (CNN) method to help with classification problems and several image processing methods. This program will be run through an android application. The results of CNN model training accuracy are 97.92%. Thus, it can be concluded that the design and testing of a model to classify the maturity level of nutmeg seeds using artificial intelligence and the implementation of the model into an android application has been successfully carried out.","PeriodicalId":507934,"journal":{"name":"IAES International Journal of Artificial Intelligence (IJ-AI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141229328","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 : 2024-06-01DOI: 10.11591/ijai.v13.i2.pp1868-1877
Liliana Swastina, B. Rahmatullah, Aslina Saad, Hussin Khan
The monitoring of children's nutritional status serves as a crucial tool for assessing the health of both children and society as a whole. In this regard, machine learning has been employed to predict nutritional status for monitoring purposes. This topic has been extensively discussed; however, the question remains as to which algorithm or machine learning framework can yield the highest accuracy in predicting the nutritional status of children within a specific region. Furthermore, determining the appropriate dataset for predictions is also crucial. Therefore, this review aims to identify and analyze the research trends, dataset characteristics, algorithms, and frameworks utilized in studies pertaining to the nutritional status of children under the age of five from 2017 to early 2022. The selected papers focus on the application of machine learning techniques in predicting nutritional status. The findings of this research reveal that the Bangladesh DHS 2014 dataset is among the popular choices for machine learning applications in this field. The most commonly employed algorithms include Neural Networks, Random Forests, Logistic Regression, and Decision Trees which demonstrated promising performance. Lastly, the data preprocessing stage within a framework plays a significant role in models aimed at predicting nutritional status.
{"title":"A systematic review on research trends, datasets, algorithms, and frameworks of children’s nutritional status prediction","authors":"Liliana Swastina, B. Rahmatullah, Aslina Saad, Hussin Khan","doi":"10.11591/ijai.v13.i2.pp1868-1877","DOIUrl":"https://doi.org/10.11591/ijai.v13.i2.pp1868-1877","url":null,"abstract":"The monitoring of children's nutritional status serves as a crucial tool for assessing the health of both children and society as a whole. In this regard, machine learning has been employed to predict nutritional status for monitoring purposes. This topic has been extensively discussed; however, the question remains as to which algorithm or machine learning framework can yield the highest accuracy in predicting the nutritional status of children within a specific region. Furthermore, determining the appropriate dataset for predictions is also crucial. Therefore, this review aims to identify and analyze the research trends, dataset characteristics, algorithms, and frameworks utilized in studies pertaining to the nutritional status of children under the age of five from 2017 to early 2022. The selected papers focus on the application of machine learning techniques in predicting nutritional status. The findings of this research reveal that the Bangladesh DHS 2014 dataset is among the popular choices for machine learning applications in this field. The most commonly employed algorithms include Neural Networks, Random Forests, Logistic Regression, and Decision Trees which demonstrated promising performance. Lastly, the data preprocessing stage within a framework plays a significant role in models aimed at predicting nutritional status.","PeriodicalId":507934,"journal":{"name":"IAES International Journal of Artificial Intelligence (IJ-AI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141230488","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 : 2024-06-01DOI: 10.11591/ijai.v13.i2.pp1250-1261
Abraham Rajan, Manohar Manur
Sentiment analysis is the task of analysing, processing, inferencing and concluding the subjective texts along with sentiment. Considering the application of sentiment analysis, it is categorized into document-level, sentence-level and aspect level. In past, several researches have achieved solutions through the bidirectional encoder representations from transformers (BERT) model, however, the existing model does not understand the context of the aspect in deep, which leads to low metrics. This research work leads to the study of the aspect-based sentiment analysis presented by deep context bidirectional encoder representations from transformers (DC-BERT), main aim of the DC-BERT model is to improvise the context understating for aspects to enhance the metrics. DC-BERT model comprises fine-tuned BERT model along with a deep context features layer, which enables the model to understand the context of targeted aspects deeply. A customized feature layer is introduced to extract two distinctive features, later both features are integrated through the interaction layer. DC-BERT mode is evaluated considering the review dataset of laptops and restaurants from SemEval 2014 task 4, evaluation is carried out considering the different metrics. In comparison with the other model, DC-BERT achieves an accuracy of 84.48% and 92.86% for laptop and restaurant datasets respectively.
{"title":"Aspect based sentiment analysis using fine-tuned BERT model with deep context features","authors":"Abraham Rajan, Manohar Manur","doi":"10.11591/ijai.v13.i2.pp1250-1261","DOIUrl":"https://doi.org/10.11591/ijai.v13.i2.pp1250-1261","url":null,"abstract":"Sentiment analysis is the task of analysing, processing, inferencing and concluding the subjective texts along with sentiment. Considering the application of sentiment analysis, it is categorized into document-level, sentence-level and aspect level. In past, several researches have achieved solutions through the bidirectional encoder representations from transformers (BERT) model, however, the existing model does not understand the context of the aspect in deep, which leads to low metrics. This research work leads to the study of the aspect-based sentiment analysis presented by deep context bidirectional encoder representations from transformers (DC-BERT), main aim of the DC-BERT model is to improvise the context understating for aspects to enhance the metrics. DC-BERT model comprises fine-tuned BERT model along with a deep context features layer, which enables the model to understand the context of targeted aspects deeply. A customized feature layer is introduced to extract two distinctive features, later both features are integrated through the interaction layer. DC-BERT mode is evaluated considering the review dataset of laptops and restaurants from SemEval 2014 task 4, evaluation is carried out considering the different metrics. In comparison with the other model, DC-BERT achieves an accuracy of 84.48% and 92.86% for laptop and restaurant datasets respectively.","PeriodicalId":507934,"journal":{"name":"IAES International Journal of Artificial Intelligence (IJ-AI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141230511","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 : 2024-06-01DOI: 10.11591/ijai.v13.i2.pp1326-1338
Kouame Yann Olivier Akansie, Rajashekhar C. Biradar, R. Karthik, Geetha D. Devanagavi
Mobile robots are used in a variety of applications including research, education, healthcare, customer service, security and so on. Based upon the application, the robots employ different locomotion systems for their mobility. When it comes to rolling locomotion, the wheels used to provide mobility to robots can be categorized as: tracks, omnidirectional wheels, and unidirectional wheels with a steering system. The ability of omnidirectional wheels to drive machines in small spaces makes them interesting to use. Among the types of omnidirectional wheels, mecanum wheels are widely used due to their inherent benefits. With the right control strategy, robots equipped with mecanum wheels can move freely, in all possible directions. In this study, a triangle decomposition approach is employed for controlling omnidirectional mecanum wheel-based robots. The method consists of breaking down any path into a set of linear motions that can be horizontal, vertical, or oblique. Furthermore, the oblique paths are divided into smaller segments that can be resolved into a horizontal and vertical component in a right-angle triangle. The suggested control method is tested and proved on a simple scenario using Webots simulation software.
{"title":"A triangle decomposition method for the mobility control of mecanum wheel-based robots","authors":"Kouame Yann Olivier Akansie, Rajashekhar C. Biradar, R. Karthik, Geetha D. Devanagavi","doi":"10.11591/ijai.v13.i2.pp1326-1338","DOIUrl":"https://doi.org/10.11591/ijai.v13.i2.pp1326-1338","url":null,"abstract":"Mobile robots are used in a variety of applications including research, education, healthcare, customer service, security and so on. Based upon the application, the robots employ different locomotion systems for their mobility. When it comes to rolling locomotion, the wheels used to provide mobility to robots can be categorized as: tracks, omnidirectional wheels, and unidirectional wheels with a steering system. The ability of omnidirectional wheels to drive machines in small spaces makes them interesting to use. Among the types of omnidirectional wheels, mecanum wheels are widely used due to their inherent benefits. With the right control strategy, robots equipped with mecanum wheels can move freely, in all possible directions. In this study, a triangle decomposition approach is employed for controlling omnidirectional mecanum wheel-based robots. The method consists of breaking down any path into a set of linear motions that can be horizontal, vertical, or oblique. Furthermore, the oblique paths are divided into smaller segments that can be resolved into a horizontal and vertical component in a right-angle triangle. The suggested control method is tested and proved on a simple scenario using Webots simulation software.","PeriodicalId":507934,"journal":{"name":"IAES International Journal of Artificial Intelligence (IJ-AI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141232344","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 : 2024-06-01DOI: 10.11591/ijai.v13.i2.pp1858-1867
Kar Yan Choong, S. Sudin, Rafikha Aliana A. Raof, Rhui Jaan Ong
The significance of the agriculture sector in Malaysia is often overlooked, and there is a notable deficiency in the advancement of digitalization within the country's agricultural domain. The integration of a price forecasting model in the platform enables the relevant parties, including farmers, to make informed decisions and plan their crop selection based on projected future prices. In this research, the authors proposed the hybrid approach with the combination of linear model and non-linear model in doing the vegetable price forecasting model. The hybrid SARIMA-DWT-GANN model is utilized to forecast the monthly vegetable prices in Malaysia. The historical vegetable price data is collected from the FAMA Malaysia and split into training/test set for modelling. The performance of the models is evaluated on the accuracy metrics including MAE, MAPE, and RMSE. The forecasted results using the proposed hybrid model are compared to that using the single SARIMA model. In conclusion, the hybrid SARIMA-DWT-GANN model is superior to the individual model, which obtained the smaller MAE, RMSE, and got the forecast accuracy of at least 95%.
{"title":"Hybrid approach for vegetable price forecasting in electronic commerce platform","authors":"Kar Yan Choong, S. Sudin, Rafikha Aliana A. Raof, Rhui Jaan Ong","doi":"10.11591/ijai.v13.i2.pp1858-1867","DOIUrl":"https://doi.org/10.11591/ijai.v13.i2.pp1858-1867","url":null,"abstract":"The significance of the agriculture sector in Malaysia is often overlooked, and there is a notable deficiency in the advancement of digitalization within the country's agricultural domain. The integration of a price forecasting model in the platform enables the relevant parties, including farmers, to make informed decisions and plan their crop selection based on projected future prices. In this research, the authors proposed the hybrid approach with the combination of linear model and non-linear model in doing the vegetable price forecasting model. The hybrid SARIMA-DWT-GANN model is utilized to forecast the monthly vegetable prices in Malaysia. The historical vegetable price data is collected from the FAMA Malaysia and split into training/test set for modelling. The performance of the models is evaluated on the accuracy metrics including MAE, MAPE, and RMSE. The forecasted results using the proposed hybrid model are compared to that using the single SARIMA model. In conclusion, the hybrid SARIMA-DWT-GANN model is superior to the individual model, which obtained the smaller MAE, RMSE, and got the forecast accuracy of at least 95%. ","PeriodicalId":507934,"journal":{"name":"IAES International Journal of Artificial Intelligence (IJ-AI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141233723","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 : 2024-06-01DOI: 10.11591/ijai.v13.i2.pp2342-2353
Bowonsak Srisungsittisunti, Jirawat Duangkaew, S. Mekruksavanich, Nakarin Chaikaew, P. Rojanavasu
The use of JavaScript Object Notation (JSON) format as a Not only Structured Query Language (NoSQL) storage solution has grown in popularity, but has presented technical challenges, particularly in indexing large-scale JSON files. This has resulted in slow data retrieval, especially for larger datasets. In this study, we propose the use of JSON datasets to preserve data in resource survey processes. We conducted experiments on a 32-gigabyte dataset containing 1,000,000 transactions in JSON format and implemented two indexing methods, dense and sparse, to improve retrieval efficiency. Additionally, we determined the optimal range of segment sizes for the indexing methods. Our findings revealed that adopting dense indexing reduced data retrieval time from 15,635 milliseconds to 55 milliseconds in one-to-one data retrieval, and from 38,300 milliseconds to 1 millisecond in the absence of keywords. In contrast, using sparse indexing reduced data retrieval time from 33,726 milliseconds to 36 milliseconds in one-to-many data retrieval and from 47,203 milliseconds to 0.17 milliseconds when keywords were not found. Furthermore, we discovered that the optimal segment size range was between 20,000 and 200,000 transactions for both dense and sparse indexing.
{"title":"Enhancing data retrieval efficiency in large-scale JavaScript object notation datasets by using indexing techniques","authors":"Bowonsak Srisungsittisunti, Jirawat Duangkaew, S. Mekruksavanich, Nakarin Chaikaew, P. Rojanavasu","doi":"10.11591/ijai.v13.i2.pp2342-2353","DOIUrl":"https://doi.org/10.11591/ijai.v13.i2.pp2342-2353","url":null,"abstract":"<div class=\"translate-tooltip-mtz green sm-root translate hidden_translate\"><div class=\"header-wrapper\"><div class=\"header-controls\"><span>The use of JavaScript Object Notation (JSON) format as a Not only Structured Query Language (NoSQL) storage solution has grown in popularity, but has presented technical challenges, particularly in indexing large-scale JSON files. This has resulted in slow data retrieval, especially for larger datasets. In this study, we propose the use of JSON datasets to preserve data in resource survey processes. We conducted experiments on a 32-gigabyte dataset containing 1,000,000 transactions in JSON format and implemented two indexing methods, dense and sparse, to improve retrieval efficiency. Additionally, we determined the optimal range of segment sizes for the indexing methods. Our findings revealed that adopting dense indexing reduced data retrieval time from 15,635 milliseconds to 55 milliseconds in one-to-one data retrieval, and from 38,300 milliseconds to 1 millisecond in the absence of keywords. In contrast, using sparse indexing reduced data retrieval time from 33,726 milliseconds to 36 milliseconds in one-to-many data retrieval and from 47,203 milliseconds to 0.17 milliseconds when keywords were not found. Furthermore, we discovered that the optimal segment size range was between 20,000 and 200,000 transactions for both dense and sparse indexing.</span></div></div></div>","PeriodicalId":507934,"journal":{"name":"IAES International Journal of Artificial Intelligence (IJ-AI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141233964","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 : 2024-06-01DOI: 10.11591/ijai.v13.i2.pp1272-1281
Bettadapura A. Sujathakumari, Sudarshan Patil Kulkarni, Vikas Hallikeri
Alzheimer's is a progressive neurodegenerative disorder and is considered the sixth leading cause of death after cancer and heart attack. Early detection and diagnosis provide individuals to go through a wider variety of clinical trials and get multiple medical benefits. Research on the application of deep learning and machine learning to the early detection of Alzheimer's disease has recently gained considerable attention. In this paper, we propose a deep learning classification framework to classify the individual with different progression stages of Alzheimer's disease such as mild cognitive impairment (MCI) and cognitive normal (CN). The dataset from Alzheimer’s disease neuroimaging initiative (ADNI) is considered in this paper which is a multisite having collection of Neuroimaging data for researchers. Structural magnetic resonance imaging (MRI) images are considered from the ADNI data set and feature extraction is done using a 2D discrete wavelet transform. 97% of data reduction is achieved during data pre-processing. The algorithm is trained and validated. The algorithm is accelerated in Nvidia Tx2 graphics processing unit (GPU) to get the better throughput. The result shows our algorithm outperforms the other deep learning algorithms with 91.56% accuracy.
{"title":"Brain magnetic resonance imaging image classification for Alzheimer's disease and its hardware acceleration","authors":"Bettadapura A. Sujathakumari, Sudarshan Patil Kulkarni, Vikas Hallikeri","doi":"10.11591/ijai.v13.i2.pp1272-1281","DOIUrl":"https://doi.org/10.11591/ijai.v13.i2.pp1272-1281","url":null,"abstract":"Alzheimer's is a progressive neurodegenerative disorder and is considered the sixth leading cause of death after cancer and heart attack. Early detection and diagnosis provide individuals to go through a wider variety of clinical trials and get multiple medical benefits. Research on the application of deep learning and machine learning to the early detection of Alzheimer's disease has recently gained considerable attention. In this paper, we propose a deep learning classification framework to classify the individual with different progression stages of Alzheimer's disease such as mild cognitive impairment (MCI) and cognitive normal (CN). The dataset from Alzheimer’s disease neuroimaging initiative (ADNI) is considered in this paper which is a multisite having collection of Neuroimaging data for researchers. Structural magnetic resonance imaging (MRI) images are considered from the ADNI data set and feature extraction is done using a 2D discrete wavelet transform. 97% of data reduction is achieved during data pre-processing. The algorithm is trained and validated. The algorithm is accelerated in Nvidia Tx2 graphics processing unit (GPU) to get the better throughput. The result shows our algorithm outperforms the other deep learning algorithms with 91.56% accuracy. ","PeriodicalId":507934,"journal":{"name":"IAES International Journal of Artificial Intelligence (IJ-AI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141234378","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 : 2024-06-01DOI: 10.11591/ijai.v13.i2.pp1469-1478
Elia Erwani Hassan, Hanan Izzati Mohd Noor, Mohd Ruzaini Bin Hashim, M. F. Sulaima, N. Bahaman
Today, most power plants worldwide use fossil fuels such as natural gas, coal, and oil as the primary resource for energy reproduction primarily. The new term for economic environmental power dispatch (EEPD) problems is on the minimum total cost of the generator and fossil fuel emissions to address atmosphere pollution. Thus, the significant objective functions are identified to minimize the cost of generation, most minor emission pollutants, and lowest system losses individually. As an alternative, an Artificial Bee Colony (ABC) swarming algorithm is applied to solve the EEPD problem separately in the power systems on both standard IEEE 26 bus system and IEEE 57 bus system using a MATLAB programming environment. The performance of the introduced algorithm is measured based on simple mathematical analysis such as a simple deviation and its percentage from the obtained results. From the mathematical measurement, the ABC algorithm showed an improvement on each identified single objective function as compared with the gradient approach of using the Newton Raphson method in a short computational time.
{"title":"Optimal economic environmental power dispatch by using artificial bee colony algorithm","authors":"Elia Erwani Hassan, Hanan Izzati Mohd Noor, Mohd Ruzaini Bin Hashim, M. F. Sulaima, N. Bahaman","doi":"10.11591/ijai.v13.i2.pp1469-1478","DOIUrl":"https://doi.org/10.11591/ijai.v13.i2.pp1469-1478","url":null,"abstract":"Today, most power plants worldwide use fossil fuels such as natural gas, coal, and oil as the primary resource for energy reproduction primarily. The new term for economic environmental power dispatch (EEPD) problems is on the minimum total cost of the generator and fossil fuel emissions to address atmosphere pollution. Thus, the significant objective functions are identified to minimize the cost of generation, most minor emission pollutants, and lowest system losses individually. As an alternative, an Artificial Bee Colony (ABC) swarming algorithm is applied to solve the EEPD problem separately in the power systems on both standard IEEE 26 bus system and IEEE 57 bus system using a MATLAB programming environment. The performance of the introduced algorithm is measured based on simple mathematical analysis such as a simple deviation and its percentage from the obtained results. From the mathematical measurement, the ABC algorithm showed an improvement on each identified single objective function as compared with the gradient approach of using the Newton Raphson method in a short computational time.","PeriodicalId":507934,"journal":{"name":"IAES International Journal of Artificial Intelligence (IJ-AI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141231982","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 : 2024-06-01DOI: 10.11591/ijai.v13.i2.pp1668-1678
Abraham Rajan, Manohar Manur
Aspect-based sentiment analysis (ABSA) is a fine-grained task in natural language processing, which aims to predict the sentiment polarity of several parts of a sentence or document. The essential aspect of sentiment polarity and global context have deep relationships that have not received enough attention. This research work design and develops a novel ensemble deep network (EDN) which comprises the various network and integrated to enhance the model performance. In the proposed work the words of the input sentence are converted into word vectors using the optimised bidirectional encoder representations from transformers (BERT) model and an optimised BERT-graph neural networks (GNN) model with convolutions is built that analyses the ABSA of the input sentence. The optimised GNN model with convolutions for context-based word representations is developed for the word-vector embedding. We propose a novel EDN for an ABSA model for optimised BERT over GNN with convolutions. The proposed ensemble deep network proposed system (EDN-PS) is evaluated with various existing techniques and results are plotted in terms of metrics for accuracy and F1- score, concluding that the proposed EDN-PS ensures better performance in comparison with the existing model.
{"title":"Aspect based sentiment analysis using a novel ensemble deep network","authors":"Abraham Rajan, Manohar Manur","doi":"10.11591/ijai.v13.i2.pp1668-1678","DOIUrl":"https://doi.org/10.11591/ijai.v13.i2.pp1668-1678","url":null,"abstract":"Aspect-based sentiment analysis (ABSA) is a fine-grained task in natural language processing, which aims to predict the sentiment polarity of several parts of a sentence or document. The essential aspect of sentiment polarity and global context have deep relationships that have not received enough attention. This research work design and develops a novel ensemble deep network (EDN) which comprises the various network and integrated to enhance the model performance. In the proposed work the words of the input sentence are converted into word vectors using the optimised bidirectional encoder representations from transformers (BERT) model and an optimised BERT-graph neural networks (GNN) model with convolutions is built that analyses the ABSA of the input sentence. The optimised GNN model with convolutions for context-based word representations is developed for the word-vector embedding. We propose a novel EDN for an ABSA model for optimised BERT over GNN with convolutions. The proposed ensemble deep network proposed system (EDN-PS) is evaluated with various existing techniques and results are plotted in terms of metrics for accuracy and F1- score, concluding that the proposed EDN-PS ensures better performance in comparison with the existing model.","PeriodicalId":507934,"journal":{"name":"IAES International Journal of Artificial Intelligence (IJ-AI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141229278","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 : 2024-06-01DOI: 10.11591/ijai.v13.i2.pp2081-2092
Samah W. G. AbuSalim, Nordin Zakaria, Aarish Maqsood, Abdul Saboor, Yew Kwang Hooi, Norehan Mokhtar, Said Jadid Abdulkadir
Accurate detection and classification of teeth is the first step in dental disease diagnosis. However, the same class of tooth exhibits significant variations in surface appearance. Moreover, the complex geometrical structure poses challenges in learning discriminative features among the different tooth classes. Due to these complex features, tooth classification is one of the challenging research domains in deep learning. To address the aforementioned issues, the presented study proposes discriminative local feature extraction at different granular levels using YOLO models. However, this necessitates a granular intra-oral image dataset. To facilitate this requirement, a dataset at three granular levels (two, four, and seven teeth classes) is developed. YOLOv5, YOLOv6, and YOLOv7 models were trained using 2,790 images. The results indicate superior performance of YOLOv6 for two-class classification problems. The model generated a mean average precision (mAP) value of 94%. However, as the granularity level is increased, the performance of YOLO models decreases. For, four and seven-class classification problems, the highest mAP value of 87% and 79% was achieved by YOLOv5 respectively. The results indicate that different levels of granularity play an important role in tooth detection and classification. The YOLO’s performance gradually decreased as the granularity decreased especially at the finest granular level.
{"title":"Multi-granularity tooth analysis via YOLO-based object detection models for effective tooth detection and classification","authors":"Samah W. G. AbuSalim, Nordin Zakaria, Aarish Maqsood, Abdul Saboor, Yew Kwang Hooi, Norehan Mokhtar, Said Jadid Abdulkadir","doi":"10.11591/ijai.v13.i2.pp2081-2092","DOIUrl":"https://doi.org/10.11591/ijai.v13.i2.pp2081-2092","url":null,"abstract":"Accurate detection and classification of teeth is the first step in dental disease diagnosis. However, the same class of tooth exhibits significant variations in surface appearance. Moreover, the complex geometrical structure poses challenges in learning discriminative features among the different tooth classes. Due to these complex features, tooth classification is one of the challenging research domains in deep learning. To address the aforementioned issues, the presented study proposes discriminative local feature extraction at different granular levels using YOLO models. However, this necessitates a granular intra-oral image dataset. To facilitate this requirement, a dataset at three granular levels (two, four, and seven teeth classes) is developed. YOLOv5, YOLOv6, and YOLOv7 models were trained using 2,790 images. The results indicate superior performance of YOLOv6 for two-class classification problems. The model generated a mean average precision (mAP) value of 94%. However, as the granularity level is increased, the performance of YOLO models decreases. For, four and seven-class classification problems, the highest mAP value of 87% and 79% was achieved by YOLOv5 respectively. The results indicate that different levels of granularity play an important role in tooth detection and classification. The YOLO’s performance gradually decreased as the granularity decreased especially at the finest granular level.","PeriodicalId":507934,"journal":{"name":"IAES International Journal of Artificial Intelligence (IJ-AI)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141229835","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：