A pandemic epidemic called the coronavirus (COVID-19) has already afflicted people all across the world. Radiologists can visually detect coronavirus infection using a chest X-ray. This study examines two methods for categorizing COVID-19 patients based on chest x-rays: pure deep learning and traditional machine learning. In the first model, three deep learning classifiers' decisions are combined using two distinct decision fusion strategies (majority voting and Bayes optimal). To enhance classification performance, the second model merges the ideas of decision and feature fusion. Using the fusion procedure, feature vectors from deep learning models generate a feature set. The classification metrics of conventional machine learning classifiers were then optimized using a voting classifier. The first proposed model performs better than the second model when it concerns diagnosing binary and multiclass classification. The first model obtains an AUC of 0.998 for multi-class classification and 0.9755 for binary classification. The second model obtains a binary classification AUC of 0.9563 and a multiclass classification AUC of 0.968. The suggested models perform better than both the standard learners and state-of-the-art and state-of-the-art methods.
{"title":"Ensemble deep models for covid-19 pandemic classification using chest x-ray images via different fusion techniques","authors":"Lamiaa Menshawy, Ahmad Eid, Rehab F. Abdel-Kader","doi":"10.26555/ijain.v9i1.922","DOIUrl":"https://doi.org/10.26555/ijain.v9i1.922","url":null,"abstract":"A pandemic epidemic called the coronavirus (COVID-19) has already afflicted people all across the world. Radiologists can visually detect coronavirus infection using a chest X-ray. This study examines two methods for categorizing COVID-19 patients based on chest x-rays: pure deep learning and traditional machine learning. In the first model, three deep learning classifiers' decisions are combined using two distinct decision fusion strategies (majority voting and Bayes optimal). To enhance classification performance, the second model merges the ideas of decision and feature fusion. Using the fusion procedure, feature vectors from deep learning models generate a feature set. The classification metrics of conventional machine learning classifiers were then optimized using a voting classifier. The first proposed model performs better than the second model when it concerns diagnosing binary and multiclass classification. The first model obtains an AUC of 0.998 for multi-class classification and 0.9755 for binary classification. The second model obtains a binary classification AUC of 0.9563 and a multiclass classification AUC of 0.968. The suggested models perform better than both the standard learners and state-of-the-art and state-of-the-art methods.","PeriodicalId":52195,"journal":{"name":"International Journal of Advances in Intelligent Informatics","volume":"2 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80182562","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}
Nastiti Susetyo Fanany Putri, A. Wibawa, Harits Ar Rasyid, A. Nafalski, Ummi Rabaah Hasyim
In recent years, data processing has become an issue across all disciplines. Good data processing can provide decision-making recommendations. Data processing is covered in academic data processing publications, including those in computer science. This topic has grown over the past three years, demonstrating that data processing is expanding and diversifying, and there is a great deal of interest in this area of study. Within the journal, groupings (quartiles) indicate the journal's influence on other similar studies. SCImago provides this category. There are four quartiles, with the highest quartile being 1 and the lowest being 4. There are, however, numerous differences in class quartiles, with different quartile values for the same journal in different disciplines. Therefore, a method of categorization is provided to solve this issue. Classification is a machine-learning technique that groups data based on the supplied label class. Ensemble Boosting and Bagging with Decision Tree (DT) and Gaussian Nave Bayes (GNB) were utilized in this study. Several modifications were made to the ensemble algorithm's depth and estimator settings to examine the influence of adding values on the resultant precision. In the DT algorithm, both variables are altered, whereas, in the GNB algorithm, just the estimator's value is modified. Based on the average value of the accuracy results, it is known that the best algorithm for computer science datasets is GNB Bagging, with values of 68.96%, 70.99%, and 69.05%. Second-place XGBDT has 67.75% accuracy, 67.69% precision, and 67.83 recall. The DT Bagging method placed third with 67.31 percent recall, 68.13 percent precision, and 67.30 percent accuracy. The fourth sequence is the XGBoost GNB approach, which has an accuracy of 67.07%, a precision of 68.85%, and a recall of 67.18%. The Adaboost DT technique ranks in the fifth position with an accuracy of 63.65%, a precision of 64.21 %, and a recall of 63.63 %. Adaboost GNB is the least efficient algorithm for this dataset since it only achieves 43.19 % accuracy, 48.14 % precision, and 43.2% recall. The results are still quite far from the ideal. Hence the proposed method for journal quartile inequality issues is not advised.
{"title":"Boosting and bagging classification for computer science journal","authors":"Nastiti Susetyo Fanany Putri, A. Wibawa, Harits Ar Rasyid, A. Nafalski, Ummi Rabaah Hasyim","doi":"10.26555/ijain.v9i1.985","DOIUrl":"https://doi.org/10.26555/ijain.v9i1.985","url":null,"abstract":"In recent years, data processing has become an issue across all disciplines. Good data processing can provide decision-making recommendations. Data processing is covered in academic data processing publications, including those in computer science. This topic has grown over the past three years, demonstrating that data processing is expanding and diversifying, and there is a great deal of interest in this area of study. Within the journal, groupings (quartiles) indicate the journal's influence on other similar studies. SCImago provides this category. There are four quartiles, with the highest quartile being 1 and the lowest being 4. There are, however, numerous differences in class quartiles, with different quartile values for the same journal in different disciplines. Therefore, a method of categorization is provided to solve this issue. Classification is a machine-learning technique that groups data based on the supplied label class. Ensemble Boosting and Bagging with Decision Tree (DT) and Gaussian Nave Bayes (GNB) were utilized in this study. Several modifications were made to the ensemble algorithm's depth and estimator settings to examine the influence of adding values on the resultant precision. In the DT algorithm, both variables are altered, whereas, in the GNB algorithm, just the estimator's value is modified. Based on the average value of the accuracy results, it is known that the best algorithm for computer science datasets is GNB Bagging, with values of 68.96%, 70.99%, and 69.05%. Second-place XGBDT has 67.75% accuracy, 67.69% precision, and 67.83 recall. The DT Bagging method placed third with 67.31 percent recall, 68.13 percent precision, and 67.30 percent accuracy. The fourth sequence is the XGBoost GNB approach, which has an accuracy of 67.07%, a precision of 68.85%, and a recall of 67.18%. The Adaboost DT technique ranks in the fifth position with an accuracy of 63.65%, a precision of 64.21 %, and a recall of 63.63 %. Adaboost GNB is the least efficient algorithm for this dataset since it only achieves 43.19 % accuracy, 48.14 % precision, and 43.2% recall. The results are still quite far from the ideal. Hence the proposed method for journal quartile inequality issues is not advised.","PeriodicalId":52195,"journal":{"name":"International Journal of Advances in Intelligent Informatics","volume":"26 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87478567","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}
I. Hipiny, Hamimah Ujir, A. Alias, M. Shanat, Mohamad Khairi Ishak
Video-based action quality assessment (AQA) is a non-trivial task due to the subtle visual differences between data produced by experts and non-experts. Current methods are extended from the action recognition domain where most are based on temporal pattern matching. AQA has additional requirements where order and tempo matter for rating the quality of an action. We present a novel dataset of ranked TikTok dance videos, and a pairwise AQA method for predicting which video of a same-label pair was sourced from the better dancer. Exhaustive pairings of same-label videos were randomly assigned to 100 human annotators, ultimately producing a ranked list per label category. Our method relies on a successful detection of the subject’s 2D pose inside successive query frames where the order and tempo of actions are encoded inside a produced String sequence. The detected 2D pose returns a top-matching Visual word from a Codebook to represent the current frame. Given a same-label pair, we generate a String value of concatenated Visual words for each video. By computing the edit distance score between each String value and the Gold Standard’s (i.e., the top-ranked video(s) for that label category), we declare the video with the lower score as the winner. The pairwise AQA method is implemented using two schemes, i.e., with and without text compression. Although the average precision for both schemes over 12 label categories is low, at 0.45 with text compression and 0.48 without, precision values for several label categories are comparable to past methods’ (median: 0.47, max: 0.66).
{"title":"Who danced better? ranked tiktok dance video dataset and pairwise action quality assessment method","authors":"I. Hipiny, Hamimah Ujir, A. Alias, M. Shanat, Mohamad Khairi Ishak","doi":"10.26555/ijain.v9i1.919","DOIUrl":"https://doi.org/10.26555/ijain.v9i1.919","url":null,"abstract":"Video-based action quality assessment (AQA) is a non-trivial task due to the subtle visual differences between data produced by experts and non-experts. Current methods are extended from the action recognition domain where most are based on temporal pattern matching. AQA has additional requirements where order and tempo matter for rating the quality of an action. We present a novel dataset of ranked TikTok dance videos, and a pairwise AQA method for predicting which video of a same-label pair was sourced from the better dancer. Exhaustive pairings of same-label videos were randomly assigned to 100 human annotators, ultimately producing a ranked list per label category. Our method relies on a successful detection of the subject’s 2D pose inside successive query frames where the order and tempo of actions are encoded inside a produced String sequence. The detected 2D pose returns a top-matching Visual word from a Codebook to represent the current frame. Given a same-label pair, we generate a String value of concatenated Visual words for each video. By computing the edit distance score between each String value and the Gold Standard’s (i.e., the top-ranked video(s) for that label category), we declare the video with the lower score as the winner. The pairwise AQA method is implemented using two schemes, i.e., with and without text compression. Although the average precision for both schemes over 12 label categories is low, at 0.45 with text compression and 0.48 without, precision values for several label categories are comparable to past methods’ (median: 0.47, max: 0.66).","PeriodicalId":52195,"journal":{"name":"International Journal of Advances in Intelligent Informatics","volume":"29 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88738384","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}
Soon Hwai Ing, A. Abdullah, M. Y. Mashor, Z. Mohamed-Hussein, Z. Mohamed, W. C. Ang
Coronavirus causes a global pandemic that has adversely affected public health, the economy, including every life aspect. To manage the spread, innumerable measurements are gathered. Administering vaccines is considered to be among the precautionary steps under the blueprint. Among all vaccines, the messenger ribonucleic acid (mRNA) vaccines provide notable effectiveness with minimal side effects. However, it is easily degraded and limits its application. Therefore, considering the cruciality of predicting the degradation rate of the mRNA vaccine, this prediction study is proposed. In addition, this study compared the hybridizing sequence of the hybrid model to identify its influence on prediction performance. Five models are created for exploration and prediction on the COVID-19 mRNA vaccine dataset provided by Stanford University and made accessible on the Kaggle community platform employing the two deep learning algorithms, Long Short-Term Memory (LSTM) as well as Gated Recurrent Unit (GRU). The Mean Columnwise Root Mean Square Error (MCRMSE) performance metric was utilized to assess each model’s performance. Results demonstrated that both GRU and LSTM are befitting for predicting the degradation rate of COVID-19 mRNA vaccines. Moreover, performance improvement could be achieved by performing the hybridization approach. Among Hybrid_1, Hybrid_2, and Hybrid_3, when trained with Set_1 augmented data, Hybrid_3 with the lowest training error (0.1257) and validation error (0.1324) surpassed the other two models; the same for model training with Set_2 augmented data, scoring 0.0164 and 0.0175 MCRMSE for training error and validation error, respectively. The variance in results obtained by hybrid models from experimenting claimed hybridizing sequence of algorithms in hybrid modeling should be concerned.
{"title":"Exploration of hybrid deep learning algorithms for covid-19 mrna vaccine degradation prediction system","authors":"Soon Hwai Ing, A. Abdullah, M. Y. Mashor, Z. Mohamed-Hussein, Z. Mohamed, W. C. Ang","doi":"10.26555/ijain.v8i3.950","DOIUrl":"https://doi.org/10.26555/ijain.v8i3.950","url":null,"abstract":"Coronavirus causes a global pandemic that has adversely affected public health, the economy, including every life aspect. To manage the spread, innumerable measurements are gathered. Administering vaccines is considered to be among the precautionary steps under the blueprint. Among all vaccines, the messenger ribonucleic acid (mRNA) vaccines provide notable effectiveness with minimal side effects. However, it is easily degraded and limits its application. Therefore, considering the cruciality of predicting the degradation rate of the mRNA vaccine, this prediction study is proposed. In addition, this study compared the hybridizing sequence of the hybrid model to identify its influence on prediction performance. Five models are created for exploration and prediction on the COVID-19 mRNA vaccine dataset provided by Stanford University and made accessible on the Kaggle community platform employing the two deep learning algorithms, Long Short-Term Memory (LSTM) as well as Gated Recurrent Unit (GRU). The Mean Columnwise Root Mean Square Error (MCRMSE) performance metric was utilized to assess each model’s performance. Results demonstrated that both GRU and LSTM are befitting for predicting the degradation rate of COVID-19 mRNA vaccines. Moreover, performance improvement could be achieved by performing the hybridization approach. Among Hybrid_1, Hybrid_2, and Hybrid_3, when trained with Set_1 augmented data, Hybrid_3 with the lowest training error (0.1257) and validation error (0.1324) surpassed the other two models; the same for model training with Set_2 augmented data, scoring 0.0164 and 0.0175 MCRMSE for training error and validation error, respectively. The variance in results obtained by hybrid models from experimenting claimed hybridizing sequence of algorithms in hybrid modeling should be concerned.","PeriodicalId":52195,"journal":{"name":"International Journal of Advances in Intelligent Informatics","volume":"38 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85683583","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}
Advances in information technology have given rise to online hotel reservation options. The user review feature is an important factor during the online booking of hotels. Generally, most online hotel booking service providers provide review and rating features for assessing hotels. However, not all service providers provide rating features or recap reviews for every aspect of the hotel services offered. Therefore, we propose a method to summarise reviews based on multiple aspects, including food, room, service, and location. This method uses long short-term memory (LSTM), together with hidden layers and automation of the optimal number of hidden neurons. The F1-measure value of 75.28% for the best model was based on the fact that (i) the size of the first hidden layer is 1,200 neurons with the tanh activation function, and (ii) the size of the second hidden layer is 600 neurons with the ReLU activation function. The proposed model outperforms the baseline model (also known as standard LSTM) by 10.16%. It is anticipated that the model developed through this study can be accessed by users of online hotel booking services to acquire a review recap on more specific aspects of services offered by hotels
{"title":"Aspect-based sentiment analysis for hotel reviews using an improved model of long short-term memory","authors":"Rahmat Jayanto, R. Kusumaningrum, A. Wibowo","doi":"10.26555/ijain.v8i3.691","DOIUrl":"https://doi.org/10.26555/ijain.v8i3.691","url":null,"abstract":"Advances in information technology have given rise to online hotel reservation options. The user review feature is an important factor during the online booking of hotels. Generally, most online hotel booking service providers provide review and rating features for assessing hotels. However, not all service providers provide rating features or recap reviews for every aspect of the hotel services offered. Therefore, we propose a method to summarise reviews based on multiple aspects, including food, room, service, and location. This method uses long short-term memory (LSTM), together with hidden layers and automation of the optimal number of hidden neurons. The F1-measure value of 75.28% for the best model was based on the fact that (i) the size of the first hidden layer is 1,200 neurons with the tanh activation function, and (ii) the size of the second hidden layer is 600 neurons with the ReLU activation function. The proposed model outperforms the baseline model (also known as standard LSTM) by 10.16%. It is anticipated that the model developed through this study can be accessed by users of online hotel booking services to acquire a review recap on more specific aspects of services offered by hotels","PeriodicalId":52195,"journal":{"name":"International Journal of Advances in Intelligent Informatics","volume":"7 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73678891","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}
Automated detection of threat objects in a security X-ray image is vital to prevent unwanted incidents in busy places like airports, train stations, and malls. The manual method of threat object detection is time-consuming and tedious. Also, the person on duty can overlook the threat objects due to limited time in checking every person’s belongings. As a solution, this paper presents a faster region-based convolutional neural network (Faster R-CNN) object detector to automatically identify threat objects in an X-ray image using the IEDXray dataset. The dataset was composed of scanned X-ray images of improvised explosive device (IED) replicas without the main charge. This paper extensively evaluates the Faster R-CNN architecture in threat object detection to determine which configuration can be used to improve the detection performance. Our findings showed that the proposed method could identify three classes of threat objects in X-ray images. In addition, the mean average precision (mAP) of the threat object detector could be improved by increasing the input image's image resolution but sacrificing the detector's speed. The threat object detector achieved 77.59% mAP and recorded an inference time of 208.96 ms by resizing the input image to 900 × 1536 resolution. Results also showed that increasing the bounding box proposals did not significantly improve the detection performance. The mAP using 150 bounding box proposals only achieved 75.65% mAP, and increasing the bounding box proposal twice reduced the mAP to 72.22%.
{"title":"Identifying threat objects using faster region-based convolutional neural networks (faster R-CNN)","authors":"Reagan L. Galvez, E. Dadios","doi":"10.26555/ijain.v8i3.952","DOIUrl":"https://doi.org/10.26555/ijain.v8i3.952","url":null,"abstract":"Automated detection of threat objects in a security X-ray image is vital to prevent unwanted incidents in busy places like airports, train stations, and malls. The manual method of threat object detection is time-consuming and tedious. Also, the person on duty can overlook the threat objects due to limited time in checking every person’s belongings. As a solution, this paper presents a faster region-based convolutional neural network (Faster R-CNN) object detector to automatically identify threat objects in an X-ray image using the IEDXray dataset. The dataset was composed of scanned X-ray images of improvised explosive device (IED) replicas without the main charge. This paper extensively evaluates the Faster R-CNN architecture in threat object detection to determine which configuration can be used to improve the detection performance. Our findings showed that the proposed method could identify three classes of threat objects in X-ray images. In addition, the mean average precision (mAP) of the threat object detector could be improved by increasing the input image's image resolution but sacrificing the detector's speed. The threat object detector achieved 77.59% mAP and recorded an inference time of 208.96 ms by resizing the input image to 900 × 1536 resolution. Results also showed that increasing the bounding box proposals did not significantly improve the detection performance. The mAP using 150 bounding box proposals only achieved 75.65% mAP, and increasing the bounding box proposal twice reduced the mAP to 72.22%.","PeriodicalId":52195,"journal":{"name":"International Journal of Advances in Intelligent Informatics","volume":"160 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80091107","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}
Broccoli is a plant that has many benefits. The flower parts of broccoli contain protein, calcium, vitamin A, vitamin C, and many more. However, in its cultivation, broccoli plants have obstacles such as the presence of pests and diseases that can affect production of broccoli. To avoid this, the authors build a model to identify diseases in broccoli through leaf images with a size of 128x128 pixels. The model is constructed to classify healthy leaves, and disease leaves using the image processing method that uses machine learning stages. There are several stages, including K-Means segmentation, colour feature extraction, and classification using SVM (Support Vector Machine) with RBF kernel and PSO (Particle Swarm Optimization) for reduce dimensionality data. The model that has been built compares the SVM model and the SVM-PSO model. It produces good accuracy in the training of 97.63% and testing accuracy of 94.48% for SVM-PSO and 85.82% for training, and 86.25% for testing in the SVM model. Therefore, this proposed model can produce good results in categorizing healthy and diseased leaves in broccoli.
{"title":"Broccoli leaf diseases classification using support vector machine with particle swarm optimization based on feature selection","authors":"Yulio Ferdinand, W. A. Al Maki","doi":"10.26555/ijain.v8i3.951","DOIUrl":"https://doi.org/10.26555/ijain.v8i3.951","url":null,"abstract":"Broccoli is a plant that has many benefits. The flower parts of broccoli contain protein, calcium, vitamin A, vitamin C, and many more. However, in its cultivation, broccoli plants have obstacles such as the presence of pests and diseases that can affect production of broccoli. To avoid this, the authors build a model to identify diseases in broccoli through leaf images with a size of 128x128 pixels. The model is constructed to classify healthy leaves, and disease leaves using the image processing method that uses machine learning stages. There are several stages, including K-Means segmentation, colour feature extraction, and classification using SVM (Support Vector Machine) with RBF kernel and PSO (Particle Swarm Optimization) for reduce dimensionality data. The model that has been built compares the SVM model and the SVM-PSO model. It produces good accuracy in the training of 97.63% and testing accuracy of 94.48% for SVM-PSO and 85.82% for training, and 86.25% for testing in the SVM model. Therefore, this proposed model can produce good results in categorizing healthy and diseased leaves in broccoli.","PeriodicalId":52195,"journal":{"name":"International Journal of Advances in Intelligent Informatics","volume":"17 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88480675","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}
Online social networks provide rich information that characterizes the user’s personality, his interests, hobbies, and reflects his current state. Users of social networks publish photos, posts, videos, audio, etc. every day. Online social networks (OSN) open up a wide range of research opportunities for scientists. Much research conducted in recent years using graph neural networks (GNN) has shown their advantages over conventional deep learning. In particular, the use of graph neural networks for online social network analysis seems to be the most suitable. In this article we studied the use of graph convolutional neural networks with different convolution layers (GCNConv, SAGEConv, GraphConv, GATConv, TransformerConv, GINConv) for predicting the user’s professional success in VKontakte online social network, based on data obtained from his profiles. We have used various parameters obtained from users’ personal pages in VKontakte social network (the number of friends, subscribers, interesting pages, etc.) as their features for determining the professional success, as well as networks (graphs) reflecting connections between users (followers/ friends). In this work we performed graph classification by using graph convolutional neural networks (with different types of convolution layers). The best accuracy of the graph convolutional neural network (0.88) was achieved by using the graph isomorphism network (GIN) layer. The results, obtained in this work, will serve for further studies of social success, based on metrics of personal profiles of OSN users and social graphs using neural network methods.
{"title":"Online social network user performance prediction by graph neural networks","authors":"F. Gafarov, A. Berdnikov, P. Ustin","doi":"10.26555/ijain.v8i3.859","DOIUrl":"https://doi.org/10.26555/ijain.v8i3.859","url":null,"abstract":"Online social networks provide rich information that characterizes the user’s personality, his interests, hobbies, and reflects his current state. Users of social networks publish photos, posts, videos, audio, etc. every day. Online social networks (OSN) open up a wide range of research opportunities for scientists. Much research conducted in recent years using graph neural networks (GNN) has shown their advantages over conventional deep learning. In particular, the use of graph neural networks for online social network analysis seems to be the most suitable. In this article we studied the use of graph convolutional neural networks with different convolution layers (GCNConv, SAGEConv, GraphConv, GATConv, TransformerConv, GINConv) for predicting the user’s professional success in VKontakte online social network, based on data obtained from his profiles. We have used various parameters obtained from users’ personal pages in VKontakte social network (the number of friends, subscribers, interesting pages, etc.) as their features for determining the professional success, as well as networks (graphs) reflecting connections between users (followers/ friends). In this work we performed graph classification by using graph convolutional neural networks (with different types of convolution layers). The best accuracy of the graph convolutional neural network (0.88) was achieved by using the graph isomorphism network (GIN) layer. The results, obtained in this work, will serve for further studies of social success, based on metrics of personal profiles of OSN users and social graphs using neural network methods.","PeriodicalId":52195,"journal":{"name":"International Journal of Advances in Intelligent Informatics","volume":"18 7-8","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72456621","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 cold-start problem is a condition for a new node to join a network with no available information or an isolated node. Most studies use topological network information with the Triadic Closure principles to predict links in future networks. However, the method based on the Triadic Closure principles cannot predict the future link due to no common neighbors between the predicted node pairs. Adamic Adar is one of the methods based on the Triadic Closure principles. This paper proposes three methods for extending Adamic Adar based on network metrics. The main objective is to utilize the network metrics to attract the isolated node or new node to make new relationships in the future network. The proposed method is called the extended Adamic Adar index based on Degree Centrality (DCAA), Closeness Centrality (CloCAA), and Clustering Coefficient (CluCAA). Experiments were conducted by sampling 10% of the dataset as testing data. The proposed method is examined using the four real-world networks by comparing the AUC score. Finally, the experiment results show that the DCAA and CloCAA can predict up to 99% of node pairs with a cold-start problem. DCAA and CloCAA outperform the benchmark, with an AUC score of up to 0,960. This finding shows that the extended Adamic Adar index can overcome prediction failures on node pairs with cold-start problems. In addition, prediction performance is also improved compared to the original Adamic Adar. The experiment results are promising for future research due to successfully improving the prediction performance and overcoming the cold-start problem.
{"title":"Extending adamic adar for cold-start problem in link prediction based on network metrics","authors":"H. Yuliansyah, Z. Othman, Adeela Abu Bakar","doi":"10.26555/ijain.v8i3.882","DOIUrl":"https://doi.org/10.26555/ijain.v8i3.882","url":null,"abstract":"The cold-start problem is a condition for a new node to join a network with no available information or an isolated node. Most studies use topological network information with the Triadic Closure principles to predict links in future networks. However, the method based on the Triadic Closure principles cannot predict the future link due to no common neighbors between the predicted node pairs. Adamic Adar is one of the methods based on the Triadic Closure principles. This paper proposes three methods for extending Adamic Adar based on network metrics. The main objective is to utilize the network metrics to attract the isolated node or new node to make new relationships in the future network. The proposed method is called the extended Adamic Adar index based on Degree Centrality (DCAA), Closeness Centrality (CloCAA), and Clustering Coefficient (CluCAA). Experiments were conducted by sampling 10% of the dataset as testing data. The proposed method is examined using the four real-world networks by comparing the AUC score. Finally, the experiment results show that the DCAA and CloCAA can predict up to 99% of node pairs with a cold-start problem. DCAA and CloCAA outperform the benchmark, with an AUC score of up to 0,960. This finding shows that the extended Adamic Adar index can overcome prediction failures on node pairs with cold-start problems. In addition, prediction performance is also improved compared to the original Adamic Adar. The experiment results are promising for future research due to successfully improving the prediction performance and overcoming the cold-start problem.","PeriodicalId":52195,"journal":{"name":"International Journal of Advances in Intelligent Informatics","volume":"79 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82408741","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}
Kartika Firdausy, O. Wahyunggoro, H. A. Nugroho, M. B. Sasongko
The eye-fundus photograph is widely used for eye examinations. Accurate identification of retinal blood vessels could reveal information that is helpful for clinical diagnoses of many health disorders. Although several methods have been proposed to segment images of retinal blood vessels, the sensitivity of these methods is plausible to be improved. The algorithm’s sensitivity refers to the algorithm’s ability to identify retinal vessel pixels correctly. Furthermore, the resolution and quality of retinal images are improving rapidly. Consequently, new segmentation methods are in demand to overcome issues from high-resolution images. This study presented improved performance of retinal vessel segmentation using a novel edge detection scheme based on the phase stretch transform (PST) function as its kernel. Before applying the edge detection stage, the input retinal images were pre-processed. During the pre-processing step, non-local means filtering on the green channel image, followed by contrast limited adaptive histogram equalization (CLAHE) and median filtering, were applied to enhance the retinal image. After applying the edge detection stage, the post-processing steps, including the CLAHE, median filtering, thresholding, morphological opening, and closing, were implemented to obtain the segmented image. The proposed method was evaluated using images from the high-resolution fundus (HRF) public database and yielded promising results for sensitivity improvement of retinal blood vessel detection. The proposed approach contributes to a better segmentation performance with an average sensitivity of 0.813, representing a clear improvement over several benchmark techniques
{"title":"A new approach for sensitivity improvement of retinal blood vessel segmentation in high-resolution fundus images based on phase stretch transform","authors":"Kartika Firdausy, O. Wahyunggoro, H. A. Nugroho, M. B. Sasongko","doi":"10.26555/ijain.v8i3.914","DOIUrl":"https://doi.org/10.26555/ijain.v8i3.914","url":null,"abstract":"The eye-fundus photograph is widely used for eye examinations. Accurate identification of retinal blood vessels could reveal information that is helpful for clinical diagnoses of many health disorders. Although several methods have been proposed to segment images of retinal blood vessels, the sensitivity of these methods is plausible to be improved. The algorithm’s sensitivity refers to the algorithm’s ability to identify retinal vessel pixels correctly. Furthermore, the resolution and quality of retinal images are improving rapidly. Consequently, new segmentation methods are in demand to overcome issues from high-resolution images. This study presented improved performance of retinal vessel segmentation using a novel edge detection scheme based on the phase stretch transform (PST) function as its kernel. Before applying the edge detection stage, the input retinal images were pre-processed. During the pre-processing step, non-local means filtering on the green channel image, followed by contrast limited adaptive histogram equalization (CLAHE) and median filtering, were applied to enhance the retinal image. After applying the edge detection stage, the post-processing steps, including the CLAHE, median filtering, thresholding, morphological opening, and closing, were implemented to obtain the segmented image. The proposed method was evaluated using images from the high-resolution fundus (HRF) public database and yielded promising results for sensitivity improvement of retinal blood vessel detection. The proposed approach contributes to a better segmentation performance with an average sensitivity of 0.813, representing a clear improvement over several benchmark techniques","PeriodicalId":52195,"journal":{"name":"International Journal of Advances in Intelligent Informatics","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76118801","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：