Pub Date : 2019-10-01DOI: 10.1109/ICICoS48119.2019.8982484
Muhammad Ammarul Hilmy, P. S. Sasongko
Skin cancer is a public health problem with more than 123,000 new cases diagnosed worldwide every year. System skin cancer screening reliable automatic will provide a great help for doctors to detect skin lesions as early as possible. The efficiency of deep learning based methods has recently outperformed conventional image processing methods in terms of classification. This study applied an ensemble of CNN to classify 7 categories of skin lesions. The preprocessing stage is hair removal, image resizing, and image augmentation. Model evaluation results with 1,440 test data indicate that the ensemble model achieve the best accuracy of 91.7% with a combination of learning rate parameters of le-3 and the use of dropouts in the model architecture.
{"title":"Ensembles of Convolutional Neural Networks for Skin Lesion Dermoscopy Images Classification","authors":"Muhammad Ammarul Hilmy, P. S. Sasongko","doi":"10.1109/ICICoS48119.2019.8982484","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982484","url":null,"abstract":"Skin cancer is a public health problem with more than 123,000 new cases diagnosed worldwide every year. System skin cancer screening reliable automatic will provide a great help for doctors to detect skin lesions as early as possible. The efficiency of deep learning based methods has recently outperformed conventional image processing methods in terms of classification. This study applied an ensemble of CNN to classify 7 categories of skin lesions. The preprocessing stage is hair removal, image resizing, and image augmentation. Model evaluation results with 1,440 test data indicate that the ensemble model achieve the best accuracy of 91.7% with a combination of learning rate parameters of le-3 and the use of dropouts in the model architecture.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125538678","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 : 2019-10-01DOI: 10.1109/ICICoS48119.2019.8982393
A. Sofwan, Imam Santoso, Himawan Pradipta, M. Arfan, Ajub Ajulian Zahra M
Heart rate sounds have a special pattern that is in accordance with a person's heart condition. An abnormal heart will cause a distinctive sound called a murmur. Murmurs caused by various things that indicate a person's condition. Through a Phonocardiogram (PCG), it can be seen a person's heart rate signal wave. Normal heartbeat and murmurs have a distinctive pattern, so that through this pattern it can be detected a person's heart defects. This study will make a classification program that will sense normal heart sounds and murmurs. This program uses feature extraction methods using LPC (Linear Predictive Coding) and classification using k-NN (k-Nearest Neighbor) to identify these 2 heart conditions. The data that will be used as a database consists of samples of normal heart rate sounds and murmurs, and also data obtained from the heart rate detection device in the. wav, mono format. The system for detecting heart abnormalities consists of three main parts, namely: recording heart rate sounds, feature extraction using LPC with order 10, and feature lines using k-NN with 3 types of distances and variations of k. From the results of testing with these types of distance, the obtained average accuracy value of Chebyshev, City Block, and Euclidean are 96.67, 91.67, and 93.33 percent, respectively. In addition, the value of k equal 3 is the most optimal value of k with an average level of 96.67 percent.
{"title":"Normal and Murmur Heart Sound Classification Using Linear Predictive Coding and k-Nearest Neighbor Methods","authors":"A. Sofwan, Imam Santoso, Himawan Pradipta, M. Arfan, Ajub Ajulian Zahra M","doi":"10.1109/ICICoS48119.2019.8982393","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982393","url":null,"abstract":"Heart rate sounds have a special pattern that is in accordance with a person's heart condition. An abnormal heart will cause a distinctive sound called a murmur. Murmurs caused by various things that indicate a person's condition. Through a Phonocardiogram (PCG), it can be seen a person's heart rate signal wave. Normal heartbeat and murmurs have a distinctive pattern, so that through this pattern it can be detected a person's heart defects. This study will make a classification program that will sense normal heart sounds and murmurs. This program uses feature extraction methods using LPC (Linear Predictive Coding) and classification using k-NN (k-Nearest Neighbor) to identify these 2 heart conditions. The data that will be used as a database consists of samples of normal heart rate sounds and murmurs, and also data obtained from the heart rate detection device in the. wav, mono format. The system for detecting heart abnormalities consists of three main parts, namely: recording heart rate sounds, feature extraction using LPC with order 10, and feature lines using k-NN with 3 types of distances and variations of k. From the results of testing with these types of distance, the obtained average accuracy value of Chebyshev, City Block, and Euclidean are 96.67, 91.67, and 93.33 percent, respectively. In addition, the value of k equal 3 is the most optimal value of k with an average level of 96.67 percent.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126897927","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 : 2019-10-01DOI: 10.1109/ICICoS48119.2019.8982437
G. P. Kuntarto, Yossy Alrin, I. Gunawan
The Dwipa Ontology III is a one of many sources of knowledge about the tourism domain in Indonesia. It stores information and its relations mainly about accommodation, attraction, activity, and amenity. However, this particular ontology is lack to present information about weather which highly needed by tourists during traveling to the tourism destinations. Firstly, this research aims to align two sources of weather ontologies: the weather ontology and the weather ontology for smart city in order to create a new concept of weather ontology. This new weather ontology concept is generated by measuring its linguistic similarity. Secondly, this research focus on enriching the concept of initial ontology named the Dwipa Ontology III with the concept generated by ontology alignment approaches. The ontology alignment method has been successfully mapped two weather ontologies into a new concept of weather that consist of 14 literal data: WeatherCondition, WeatherReport, dewPointAtmospere, unit, gustingWind, Celsius, Humidity, AtmosperePressure, Speed, Temperature, Interval, hasDate, hasSpeedValue, and hasUnit. This new concept has been enriched to the Dwipa Ontology III by adding one new concept/ class named weatherReport that consist of four sub classes: Unit, Interval, WeatherCondition and Temperature and a total of 255 instances. This latest version of aligned and enriched ontology is given name as Dwipa Ontology III+.
{"title":"The Key Role of Ontology Alignment and Enrichment Methodologies for Aligning and Enriching Dwipa Ontology with the Weather Concept on the Tourism Domain","authors":"G. P. Kuntarto, Yossy Alrin, I. Gunawan","doi":"10.1109/ICICoS48119.2019.8982437","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982437","url":null,"abstract":"The Dwipa Ontology III is a one of many sources of knowledge about the tourism domain in Indonesia. It stores information and its relations mainly about accommodation, attraction, activity, and amenity. However, this particular ontology is lack to present information about weather which highly needed by tourists during traveling to the tourism destinations. Firstly, this research aims to align two sources of weather ontologies: the weather ontology and the weather ontology for smart city in order to create a new concept of weather ontology. This new weather ontology concept is generated by measuring its linguistic similarity. Secondly, this research focus on enriching the concept of initial ontology named the Dwipa Ontology III with the concept generated by ontology alignment approaches. The ontology alignment method has been successfully mapped two weather ontologies into a new concept of weather that consist of 14 literal data: WeatherCondition, WeatherReport, dewPointAtmospere, unit, gustingWind, Celsius, Humidity, AtmosperePressure, Speed, Temperature, Interval, hasDate, hasSpeedValue, and hasUnit. This new concept has been enriched to the Dwipa Ontology III by adding one new concept/ class named weatherReport that consist of four sub classes: Unit, Interval, WeatherCondition and Temperature and a total of 255 instances. This latest version of aligned and enriched ontology is given name as Dwipa Ontology III+.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"761 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122995033","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 : 2019-10-01DOI: 10.1109/ICICoS48119.2019.8982512
R. Wihandika
Diabetic retinopathy (DR) is a condition which affects the eye caused by the rise of glucose in the blood. It is the primary cause of sight loss. Blood vessel is among the retinal objects which is altered by DR. By monitoring the the changes of the retinal blood vessel, severe DR or even vision loss can be avoided. Monitoring the condition of the blood vessel can be performed only by segmenting the blood vessel area from a digital fundus image. However, manual segmentation of retinal blood vessel is tedious and time-consuming, especially when processing a large number of images. Thus, automatic retinal blood vessel segmentation method is urgently required. Additionally, automatic retinal blood vessel segmentation methods are also helpful for retina-based person authentication systems. There exist various blood vessel segmentation methods. This study proposes an improved version of the line operator method based on the previous line method [1]. The proposed method is evaluated on the DRIVE dataset and shows improvement in terms of accuracy over previous methods, resulting in 96.24 % accuracy.
{"title":"Improved Line Operator for Retinal Blood Vessel Segmentation","authors":"R. Wihandika","doi":"10.1109/ICICoS48119.2019.8982512","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982512","url":null,"abstract":"Diabetic retinopathy (DR) is a condition which affects the eye caused by the rise of glucose in the blood. It is the primary cause of sight loss. Blood vessel is among the retinal objects which is altered by DR. By monitoring the the changes of the retinal blood vessel, severe DR or even vision loss can be avoided. Monitoring the condition of the blood vessel can be performed only by segmenting the blood vessel area from a digital fundus image. However, manual segmentation of retinal blood vessel is tedious and time-consuming, especially when processing a large number of images. Thus, automatic retinal blood vessel segmentation method is urgently required. Additionally, automatic retinal blood vessel segmentation methods are also helpful for retina-based person authentication systems. There exist various blood vessel segmentation methods. This study proposes an improved version of the line operator method based on the previous line method [1]. The proposed method is evaluated on the DRIVE dataset and shows improvement in terms of accuracy over previous methods, resulting in 96.24 % accuracy.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116240501","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 : 2019-10-01DOI: 10.1109/ICICoS48119.2019.8982429
Muhammad Radifan Aldiansyah, P. S. Sasongko
Sentiment Analysis was a process for identifying whether a source of text contains certain opinions, emotions, and polarity. Twitter Sentiment Analysis was a process for identifying sentiment and polarity on tweet. Twitter Sentiment Analysis provided a way for did survey about public sentiment to product, or particular even through collections of tweet. Main problem in sentiment identifying was how to determine classification model that gave high accuracy to classifying sentiment of tweet. One of the method for classifying sentiment of tweet was Deep Learning. Convolutional Neural Network (CNN) was special type of architecture from Deep Learning that its architecture had convolution layer. Convolution layer was important for extract relevant feature from text for classifying sentiment. The objective of this research was for found out the best CNN model for classifying sentiment of tweet. By using a dataset of tweets about public opinion on the Smartfren 4G network service, we searched the best CNN model using 6 combination parameters, that is the computational eficiency method, window size, and dimension of word embedding for parameters in Word2Vec Skip-gram model, then activation function in convolution layer, dropout rate, and pool size in pooling layer for parameters in CNN. The test is done using 10-fold cross validation for each search for the best parameter value and produced the best CNN model with an accuracy value of 88,21%.
{"title":"Twitter Sentiment Analysis About Public Opinion on 4G Smartfren Network Services Using Convolutional Neural Network","authors":"Muhammad Radifan Aldiansyah, P. S. Sasongko","doi":"10.1109/ICICoS48119.2019.8982429","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982429","url":null,"abstract":"Sentiment Analysis was a process for identifying whether a source of text contains certain opinions, emotions, and polarity. Twitter Sentiment Analysis was a process for identifying sentiment and polarity on tweet. Twitter Sentiment Analysis provided a way for did survey about public sentiment to product, or particular even through collections of tweet. Main problem in sentiment identifying was how to determine classification model that gave high accuracy to classifying sentiment of tweet. One of the method for classifying sentiment of tweet was Deep Learning. Convolutional Neural Network (CNN) was special type of architecture from Deep Learning that its architecture had convolution layer. Convolution layer was important for extract relevant feature from text for classifying sentiment. The objective of this research was for found out the best CNN model for classifying sentiment of tweet. By using a dataset of tweets about public opinion on the Smartfren 4G network service, we searched the best CNN model using 6 combination parameters, that is the computational eficiency method, window size, and dimension of word embedding for parameters in Word2Vec Skip-gram model, then activation function in convolution layer, dropout rate, and pool size in pooling layer for parameters in CNN. The test is done using 10-fold cross validation for each search for the best parameter value and produced the best CNN model with an accuracy value of 88,21%.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134599984","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 : 2019-10-01DOI: 10.1109/ICICoS48119.2019.8982396
A. Mulyanto, Rohmat Indra Borman, Purwono Prasetyawan, W. Jatmiko, P. Mursanto
Real-time human detecting and tracking is an important task in Advanced Driver Assistance System (ADAS) especialy in providing an information about situation in front of vehicle. Deep Convolutional Neural Networks (CNN) is one algorithm that is widely applied to classify and detect objects. CNN has shown an impressive performance. However, the high computation of Deep CNN makes the algorithm difficult to be applied to the real ADAS system. Since 2014, the One-stage Detector approach such as SSD and YOLO began to be applied on devices with low computation. In this experiment, we present a real-time system for the detection and the tracking of humans (pedestrians, cyclists, and riders) for the ADAS system implemented in Raspberry Pi 3 Model B Plus. The object detection approach in this study applies the SSD framework, and the tracking human movements approach is done by calculating the movement of midpoint coordinates from bounding box objects from two sequenced frames. The result shows the realtime human detection and tracking on Raspberry Pi 3 B devices with input frame with a height 300 and a width 300 runs at 0.8 FPS with 77.6 percent processor consumption and 70.3 percent memory. Therefore, the use of Raspberry Pi 3 B Plus for human detection and tracking in ADAS systems is not suitable for the vehicle speeds above 50 Km per hour when runs at 0.8 FPS. Then the tracking system based on the coordinate movement of the midpoint bounding box has a problem when there is a bounding box overlapping or slicing each other
在高级驾驶辅助系统(ADAS)中，人的实时检测和跟踪是一项重要的任务，特别是在提供车辆前方的情况信息方面。深度卷积神经网络(CNN)是一种广泛应用于物体分类和检测的算法。CNN的表现令人印象深刻。然而，深度CNN的高计算量使得该算法难以应用于实际的ADAS系统。从2014年开始，SSD、YOLO等单级检测器方法开始在低计算量的设备上应用。在本实验中，我们提出了一个用于检测和跟踪人类(行人，骑自行车的人和骑手)的实时系统，用于在Raspberry Pi 3 Model B Plus中实现的ADAS系统。本研究的目标检测方法采用SSD框架，跟踪人体运动方法通过计算来自两个序列帧的边界框对象的中点坐标来完成。结果表明，在输入帧高度为300、宽度为300的树莓派3b设备上，实时人体检测和跟踪以0.8 FPS的速度运行，处理器消耗77.6%，内存消耗70.3%。因此，在ADAS系统中使用树莓派3b Plus进行人体检测和跟踪不适合车速超过50 Km / h、运行速度为0.8 FPS的车辆。那么基于中点边界框坐标运动的跟踪系统就会出现边界框重叠或分割的问题
{"title":"Real-Time Human Detection and Tracking Using Two Sequential Frames for Advanced Driver Assistance System","authors":"A. Mulyanto, Rohmat Indra Borman, Purwono Prasetyawan, W. Jatmiko, P. Mursanto","doi":"10.1109/ICICoS48119.2019.8982396","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982396","url":null,"abstract":"Real-time human detecting and tracking is an important task in Advanced Driver Assistance System (ADAS) especialy in providing an information about situation in front of vehicle. Deep Convolutional Neural Networks (CNN) is one algorithm that is widely applied to classify and detect objects. CNN has shown an impressive performance. However, the high computation of Deep CNN makes the algorithm difficult to be applied to the real ADAS system. Since 2014, the One-stage Detector approach such as SSD and YOLO began to be applied on devices with low computation. In this experiment, we present a real-time system for the detection and the tracking of humans (pedestrians, cyclists, and riders) for the ADAS system implemented in Raspberry Pi 3 Model B Plus. The object detection approach in this study applies the SSD framework, and the tracking human movements approach is done by calculating the movement of midpoint coordinates from bounding box objects from two sequenced frames. The result shows the realtime human detection and tracking on Raspberry Pi 3 B devices with input frame with a height 300 and a width 300 runs at 0.8 FPS with 77.6 percent processor consumption and 70.3 percent memory. Therefore, the use of Raspberry Pi 3 B Plus for human detection and tracking in ADAS systems is not suitable for the vehicle speeds above 50 Km per hour when runs at 0.8 FPS. Then the tracking system based on the coordinate movement of the midpoint bounding box has a problem when there is a bounding box overlapping or slicing each other","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133379998","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 : 2019-10-01DOI: 10.1109/ICICoS48119.2019.8982480
Fuad Dary Rosyadi, W. Wibisono, T. Ahmad, R. Ijtihadie, Ary Mazharuddin Shidiqqi
The proliferation of the Internet of Things (IoT) devices for systems development has highlighted requirements for the devices to be able to perform various types of computations and services. They include basic computation applications that perform the simple computation to more complex and cumbersome load computation tasks. Since IoT devices are designed to be powered by battery. It has limitation in energy. This issue become one of the main challenges need to be dealt with in IoT -based application developments. Computation offloading where heavy tasks can be sent to the cloud server is one of promising technique to address this issue. However, sending large computational jobs along with the data to the cloud server not always give better results in term of energy consumptions. This paper proposes an approach to build energy-efficient computation offloading framework for an IoT-based mobile crowdsensing cluster based on the DMIPS approach. The experiments were conducted using real IoT devices and the results show that the smart offloading approach can reduce the energy consumption of the devices in performing high computation tasks compared to the full local or offloading executions.
{"title":"An Energy-Aware Computation Offloading Framework for a Mobile Crowdsensing Cluster Using DMIPS Approach","authors":"Fuad Dary Rosyadi, W. Wibisono, T. Ahmad, R. Ijtihadie, Ary Mazharuddin Shidiqqi","doi":"10.1109/ICICoS48119.2019.8982480","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982480","url":null,"abstract":"The proliferation of the Internet of Things (IoT) devices for systems development has highlighted requirements for the devices to be able to perform various types of computations and services. They include basic computation applications that perform the simple computation to more complex and cumbersome load computation tasks. Since IoT devices are designed to be powered by battery. It has limitation in energy. This issue become one of the main challenges need to be dealt with in IoT -based application developments. Computation offloading where heavy tasks can be sent to the cloud server is one of promising technique to address this issue. However, sending large computational jobs along with the data to the cloud server not always give better results in term of energy consumptions. This paper proposes an approach to build energy-efficient computation offloading framework for an IoT-based mobile crowdsensing cluster based on the DMIPS approach. The experiments were conducted using real IoT devices and the results show that the smart offloading approach can reduce the energy consumption of the devices in performing high computation tasks compared to the full local or offloading executions.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122782728","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 : 2019-10-01DOI: 10.1109/ICICoS48119.2019.8982501
Chonan Firda Odayakana Umareta, Siti Mariyah
Plagiarism is a topic of concern in the world of education. One way to overcome plagiarism is to make comparisons between documents. Due to a large number of documents, extrinsic plagiarism detection frameworks are needed to make comparisons of documents in large numbers. On the other hand, there is intelligent plagiarism in which plagiarists try to hide their actions by one of them is replacing words with semantics. Therefore, this study applies an extrinsic plagiarism detection system with a Fuzzy Semantic-Based String Similarity method which is divided into three stages, namely Preprocessing, Heuristic Retrieval (HR), and Detailed Analysis (DA). In the preprocessing stage, the removal of irrelevant characters, the division of text based on sentences, stemming, tokenization, and the elimination of stopwords were performed. The search for pairs of candidate documents in the HR stage used fingerprints and Jaccard similarity. DA stage applied fuzzy semantic based-similarity. Experiments were carried out by comparing the level of document similarity between Jaccard similarity in the HR stage and fuzzy semantic-based similarity in the DA stage because both were able to produce a level of document similarity. The results show that fuzzy semantic-based similarity is better than Jaccard similarity because it can detect semantic similarities in the form of synonyms.
{"title":"Fuzzy Semantic-Based String Similarity Experiments to Detect Plagiarism in Indonesian Documents","authors":"Chonan Firda Odayakana Umareta, Siti Mariyah","doi":"10.1109/ICICoS48119.2019.8982501","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982501","url":null,"abstract":"Plagiarism is a topic of concern in the world of education. One way to overcome plagiarism is to make comparisons between documents. Due to a large number of documents, extrinsic plagiarism detection frameworks are needed to make comparisons of documents in large numbers. On the other hand, there is intelligent plagiarism in which plagiarists try to hide their actions by one of them is replacing words with semantics. Therefore, this study applies an extrinsic plagiarism detection system with a Fuzzy Semantic-Based String Similarity method which is divided into three stages, namely Preprocessing, Heuristic Retrieval (HR), and Detailed Analysis (DA). In the preprocessing stage, the removal of irrelevant characters, the division of text based on sentences, stemming, tokenization, and the elimination of stopwords were performed. The search for pairs of candidate documents in the HR stage used fingerprints and Jaccard similarity. DA stage applied fuzzy semantic based-similarity. Experiments were carried out by comparing the level of document similarity between Jaccard similarity in the HR stage and fuzzy semantic-based similarity in the DA stage because both were able to produce a level of document similarity. The results show that fuzzy semantic-based similarity is better than Jaccard similarity because it can detect semantic similarities in the form of synonyms.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"92 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125181007","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 : 2019-10-01DOI: 10.1109/ICICoS48119.2019.8982445
R. Sigit, E. Triyana, M. Rochmad
Cataracts are the lens of the eye that becomes cloudy so that light cannot penetrate, varying according to its level from a little to total opacity. Cataracts are a leading cause of visual impairment and blindness in Indonesia and even in the world. Early detection of cataracts is known as the main setting in resisting the increase in the amount of blindness caused by cataracts. Early detection of cataracts can be seen with a slit lamp that is usually used by ophthalmologists, but the number of ophthalmologists in Indonesia is inadequate, especially in small town areas. For this reason, the researcher will make a cataract disease detection device using a smartphone. A single layer perceptron method was then used to determine the classification results in the form of normal eyes, immature cataract eyes and mature cataract eyes. Based on the results of research conducted, this cataract detection system has an accuracy of 85%.
{"title":"Cataract Detection Using Single Layer Perceptron Based on Smartphone","authors":"R. Sigit, E. Triyana, M. Rochmad","doi":"10.1109/ICICoS48119.2019.8982445","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982445","url":null,"abstract":"Cataracts are the lens of the eye that becomes cloudy so that light cannot penetrate, varying according to its level from a little to total opacity. Cataracts are a leading cause of visual impairment and blindness in Indonesia and even in the world. Early detection of cataracts is known as the main setting in resisting the increase in the amount of blindness caused by cataracts. Early detection of cataracts can be seen with a slit lamp that is usually used by ophthalmologists, but the number of ophthalmologists in Indonesia is inadequate, especially in small town areas. For this reason, the researcher will make a cataract disease detection device using a smartphone. A single layer perceptron method was then used to determine the classification results in the form of normal eyes, immature cataract eyes and mature cataract eyes. Based on the results of research conducted, this cataract detection system has an accuracy of 85%.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125232162","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 : 2019-10-01DOI: 10.1109/ICICoS48119.2019.8982385
J. Purnomo, S. Endah
Recommendation system is one of many solutions for getting information rapidly from the many data available and one of its applications is the movie recommendation system. Movie recommendation system filters information then recommends movies based on rating preferences or user information. One of the most widely used algorithms is the user based collaborative filtering algorithm (CFA) to predict movie ratings which will be recommended based on similarity between target user and other users regardless of common items or the number of movies that have been rated by both. One different approach of the CFA algorithm is a dissymmetrical percentage collaborative filtering algorithm (DSPCFA) that involves common items as a consideration of measuring similarity. This study also uses two similarity measurement methods, namely the pearson correlation similarity method and the cosine similarity method as a comparison to determine the characteristics of each measurement method. The experiment results show that the DSPCFA algorithm produces a lower error value than the CFA algorithm with an error decrease of about 5% for the RMSE evaluation method (Root-mean Squared Error) and an error decrease of about 7% using the MAE (Mean Absolute Error) evaluation method. While measurement method tested shows that the pearson correlation similarity method produces a lower error value than the cosine similarity method.
{"title":"Rating Prediction on Movie Recommendation System: Collaborative Filtering Algorithm (CFA) vs. Dissymetrical Percentage Collaborative Filtering Algorithm (DSPCFA)","authors":"J. Purnomo, S. Endah","doi":"10.1109/ICICoS48119.2019.8982385","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982385","url":null,"abstract":"Recommendation system is one of many solutions for getting information rapidly from the many data available and one of its applications is the movie recommendation system. Movie recommendation system filters information then recommends movies based on rating preferences or user information. One of the most widely used algorithms is the user based collaborative filtering algorithm (CFA) to predict movie ratings which will be recommended based on similarity between target user and other users regardless of common items or the number of movies that have been rated by both. One different approach of the CFA algorithm is a dissymmetrical percentage collaborative filtering algorithm (DSPCFA) that involves common items as a consideration of measuring similarity. This study also uses two similarity measurement methods, namely the pearson correlation similarity method and the cosine similarity method as a comparison to determine the characteristics of each measurement method. The experiment results show that the DSPCFA algorithm produces a lower error value than the CFA algorithm with an error decrease of about 5% for the RMSE evaluation method (Root-mean Squared Error) and an error decrease of about 7% using the MAE (Mean Absolute Error) evaluation method. While measurement method tested shows that the pearson correlation similarity method produces a lower error value than the cosine similarity method.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"130 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127101406","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：