Pub Date : 2019-10-01DOI: 10.1109/ICICoS48119.2019.8982498
N. Idris, Widyawan, T. B. Adji
Twitter was one of the most influential social media among users. It might give an either positive or negative impact. One of the negative impacts was the presence of radicalism content. In Indonesia radicalism was often connected to the issue of SARA (ethnicity, religion, race, and intergroup relations). It remained a public issue, requiring an analysis to process information related to radicalism. The research aimed to classify radical contents. The classification based on the types of radicalism and non-radicalism. Data were classified using LSTM. In finding higher accuracy, word2vec was used to transform words into vectors. The accuracy showed using LSTM method was compared with that obtained using SVM and k-NN. The two latest methods were the methods used by previous researchers regarding Indonesian radical contents of Twitter. Referring to the findings, LSTM showed higher accuracy 81.60%.
{"title":"Classification of Radicalism Content from Twitter Written in Indonesian Language using Long Short Term Memory","authors":"N. Idris, Widyawan, T. B. Adji","doi":"10.1109/ICICoS48119.2019.8982498","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982498","url":null,"abstract":"Twitter was one of the most influential social media among users. It might give an either positive or negative impact. One of the negative impacts was the presence of radicalism content. In Indonesia radicalism was often connected to the issue of SARA (ethnicity, religion, race, and intergroup relations). It remained a public issue, requiring an analysis to process information related to radicalism. The research aimed to classify radical contents. The classification based on the types of radicalism and non-radicalism. Data were classified using LSTM. In finding higher accuracy, word2vec was used to transform words into vectors. The accuracy showed using LSTM method was compared with that obtained using SVM and k-NN. The two latest methods were the methods used by previous researchers regarding Indonesian radical contents of Twitter. Referring to the findings, LSTM showed higher accuracy 81.60%.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"403 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":"127573827","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In Indonesia, diphtheria ranks fourth as a deadly disease after cardiovascular, tuberculosis, and pneumonia. The death rate of diphtheria is estimated to 21% with symptoms of malaise, anorexia, sore throat, and increased body temperature. The diphtheria cases which was reported in 2014 showed that East Java occupied the highest number for diphtheria cases which reached until 295, contributed to 74% cases of 22 provinces in Indonesia. In the mid-2017 until mid-2018, the Ministry of Health of the Republic of Indonesia announced that there has been an ongoing diphtheria outbreak in Indonesia. The number of diphtheria cases in East Java were highly raising up at the end of 2018. Forecasting is needed to reduce the number of diphtheria cases. The method used for forecasting is the Radial Basis Function Neural Network. Several variables are involved, including Immunization Coverage, Population Density, and Number of Cases. It is observed from the experimental results that the best model indicates only one variable involved, which is Number of Cases. This model is used to forecast the number of cases of diphtheria in Malang Regency, Surabaya City, and Sumenep Regency. The results showed that RBFNN method has a good performance for forecasting in Malang with MASE value of 0.84, Surabaya with MASE value of 0.817, and Sumenep with MASE value of 0.820, which all MASE values are less than 1.
{"title":"Diphtheria Case Number Forecasting using Radial Basis Function Neural Network","authors":"Wiwik Anggraeni, Dina Nandika, Faizal Mahananto, Yeyen Sudiarti, Cut Alna Fadhilla","doi":"10.1109/ICICoS48119.2019.8982403","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982403","url":null,"abstract":"In Indonesia, diphtheria ranks fourth as a deadly disease after cardiovascular, tuberculosis, and pneumonia. The death rate of diphtheria is estimated to 21% with symptoms of malaise, anorexia, sore throat, and increased body temperature. The diphtheria cases which was reported in 2014 showed that East Java occupied the highest number for diphtheria cases which reached until 295, contributed to 74% cases of 22 provinces in Indonesia. In the mid-2017 until mid-2018, the Ministry of Health of the Republic of Indonesia announced that there has been an ongoing diphtheria outbreak in Indonesia. The number of diphtheria cases in East Java were highly raising up at the end of 2018. Forecasting is needed to reduce the number of diphtheria cases. The method used for forecasting is the Radial Basis Function Neural Network. Several variables are involved, including Immunization Coverage, Population Density, and Number of Cases. It is observed from the experimental results that the best model indicates only one variable involved, which is Number of Cases. This model is used to forecast the number of cases of diphtheria in Malang Regency, Surabaya City, and Sumenep Regency. The results showed that RBFNN method has a good performance for forecasting in Malang with MASE value of 0.84, Surabaya with MASE value of 0.817, and Sumenep with MASE value of 0.820, which all MASE values are less than 1.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"15 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":"131011238","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/ICICoS48119.2019.8982435
M. Latief, T. Siswantining, A. Bustamam, Devvi Sarwinda
Hepatocellular carcinoma is one of the cancers that cause death in the world. We get hepatocellular carcinoma data in the form of microarray data gene expression obtained from the National Center for Biotechnology Information website consisting of 40 samples and 54675 features. The main purpose of this research is to compare the performance evaluation of Hepatocellular Carcinoma by applying feature selection to several classification algorithms. Random Forest feature selection method will be paired with several classification algorithms such as Support Vector Classification, Neural Network Classification, Random Forest, Logistic Regression, and Naïve Bayes. This study uses 5-fold cross-validation as an evaluation method. The results showed that Random Forest algorithm, Neural Network, Vector Machine Classification, and Naive Bayes show higher classification performance evaluation than without using random forest feature selection, while the Logistic Regression model provides a higher performance evaluation without using Random Forest feature selection. Support Vector Classification offers the highest performance evaluation compared to four other algorithms using feature selection, but Logistic Regression provides higher performance evaluation compared to different classification algorithms without feature selection.
{"title":"A Comparative Performance Evaluation of Random Forest Feature Selection on Classification of Hepatocellular Carcinoma Gene Expression Data","authors":"M. Latief, T. Siswantining, A. Bustamam, Devvi Sarwinda","doi":"10.1109/ICICoS48119.2019.8982435","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982435","url":null,"abstract":"Hepatocellular carcinoma is one of the cancers that cause death in the world. We get hepatocellular carcinoma data in the form of microarray data gene expression obtained from the National Center for Biotechnology Information website consisting of 40 samples and 54675 features. The main purpose of this research is to compare the performance evaluation of Hepatocellular Carcinoma by applying feature selection to several classification algorithms. Random Forest feature selection method will be paired with several classification algorithms such as Support Vector Classification, Neural Network Classification, Random Forest, Logistic Regression, and Naïve Bayes. This study uses 5-fold cross-validation as an evaluation method. The results showed that Random Forest algorithm, Neural Network, Vector Machine Classification, and Naive Bayes show higher classification performance evaluation than without using random forest feature selection, while the Logistic Regression model provides a higher performance evaluation without using Random Forest feature selection. Support Vector Classification offers the highest performance evaluation compared to four other algorithms using feature selection, but Logistic Regression provides higher performance evaluation compared to different classification algorithms without feature selection.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"11 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":"132883196","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/ICICoS48119.2019.8982485
Femilia Putri Mayranti, A. H. Saputro, W. Handayani
Pigments are a vital role in plants. Pigments can consist of several chemical structures, such as chlorophyll. Chlorophyll is a green pigment of plants can help to process photosynthetic. Chlorophyll divided into chlorophyll a and b. In this study, the authors were measured chlorophyll a and b content using hyperspectral imaging. Hyperspectral imaging had 224 full wavelengths in range 400 until 1000 nm. To measure that content, not all of 224 bands had important information of chlorophyll a and b. So that using DT method for wavelength selection had increased the performance system. The number of optimal wavelengths for chlorophyll a and b is 28 and 40 wavelengths. Comparing with several algorithms, i.e. PLSR and DT, PLSR model for full bands has the performance each chlorophyll a and b of 0.90 both for R2; also 3.25 and 3.46 for RPD. DT model for full bands has the performance each chlorophyll a and b of 0.94 and 0.96 for R2; also 4.57 and 5.02 for RPD. Then, DT with wavelength selection has improved the performance system each chlorophyll a and b of 0.99 and 0.99 for R2; also 12.00 and 13.09 for RPD.
{"title":"Chlorophyll A and B Content Measurement System of Velvet Apple Leaf in Hyperspectral Imaging","authors":"Femilia Putri Mayranti, A. H. Saputro, W. Handayani","doi":"10.1109/ICICoS48119.2019.8982485","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982485","url":null,"abstract":"Pigments are a vital role in plants. Pigments can consist of several chemical structures, such as chlorophyll. Chlorophyll is a green pigment of plants can help to process photosynthetic. Chlorophyll divided into chlorophyll a and b. In this study, the authors were measured chlorophyll a and b content using hyperspectral imaging. Hyperspectral imaging had 224 full wavelengths in range 400 until 1000 nm. To measure that content, not all of 224 bands had important information of chlorophyll a and b. So that using DT method for wavelength selection had increased the performance system. The number of optimal wavelengths for chlorophyll a and b is 28 and 40 wavelengths. Comparing with several algorithms, i.e. PLSR and DT, PLSR model for full bands has the performance each chlorophyll a and b of 0.90 both for R2; also 3.25 and 3.46 for RPD. DT model for full bands has the performance each chlorophyll a and b of 0.94 and 0.96 for R2; also 4.57 and 5.02 for RPD. Then, DT with wavelength selection has improved the performance system each chlorophyll a and b of 0.99 and 0.99 for R2; also 12.00 and 13.09 for RPD.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"14 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":"128872669","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/ICICoS48119.2019.8982458
Mutia Fadhila Putri, A. Hidayanto, E. S. Negara, Z. Abidin, P. Utari, N. Budi
Game mechanics are the most visible part of gamification that designed for interaction with the game state to produce an engaging experience. In a product with gamification, choosing the game mechanics has become the primary focus because it must be appropriate with the product goal. This study aims to investigate the most suitable game mechanics for gamification in mobile payment. AHP-TOPSIS methods used for the process of ranking and selecting the best game mechanics. The criteria and sub-criteria have been determined based on the Uses and Gratification perspective. Hedonic, utilitarian, and social gratification identified as criteria. While enjoyment, passing the time, ease of use, self-presentation, information quality, economic rewards, social value, and social interaction identified as sub-criteria. The questionnaire consists of pairwise comparison, and compatibility assessment of criteria and sub-criteria was conduct and distribute to collecting data from respondents. The results from the processes of AHP-TOPSIS identified feedback as the most suitable game mechanics for gamification in mobile payment. The consistency ratio from consistency checking is CR= 0.013514, and this is acceptable as consistent with the value of CR< 0. 1.
{"title":"Ranking of Game Mechanics for Gamification in Mobile Payment Using AHP-TOPSIS: Uses and Gratification Perspective","authors":"Mutia Fadhila Putri, A. Hidayanto, E. S. Negara, Z. Abidin, P. Utari, N. Budi","doi":"10.1109/ICICoS48119.2019.8982458","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982458","url":null,"abstract":"Game mechanics are the most visible part of gamification that designed for interaction with the game state to produce an engaging experience. In a product with gamification, choosing the game mechanics has become the primary focus because it must be appropriate with the product goal. This study aims to investigate the most suitable game mechanics for gamification in mobile payment. AHP-TOPSIS methods used for the process of ranking and selecting the best game mechanics. The criteria and sub-criteria have been determined based on the Uses and Gratification perspective. Hedonic, utilitarian, and social gratification identified as criteria. While enjoyment, passing the time, ease of use, self-presentation, information quality, economic rewards, social value, and social interaction identified as sub-criteria. The questionnaire consists of pairwise comparison, and compatibility assessment of criteria and sub-criteria was conduct and distribute to collecting data from respondents. The results from the processes of AHP-TOPSIS identified feedback as the most suitable game mechanics for gamification in mobile payment. The consistency ratio from consistency checking is CR= 0.013514, and this is acceptable as consistent with the value of CR< 0. 1.","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":"115500917","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/ICICoS48119.2019.8982522
Eko Wahyudi, R. Kusumaningrum
User ratings on products sold bye-commerce greatly influence the number of purchases. Positive ratings will encourage other buyers to participate in buying the product. While negative ratings given by users will reduce the interest in purchasing products. Nonconformities between rating and user reviews sometimes provide a wrong assessment of a product. This happens because buyers also provide reviews on the quality of delivery services from e-commerce. Based on that issue, the utilization of the Latent Dirichlet Allocation (LDA) could be used on sentiment analysis of the user reviews. Sentiment analysis of the user reviews aims to facilitate e-commerce in informing the product quality as rating supporters that have been given by users. This research aims to determine the classification performance of sentiment analysis on e-commerce user reviews using the LDA algorithm with input data in the form of e-commerce user reviews. Then, compare the application of sentiment analysis of the user reviews with the use of general training data and per category training data. The result of this research showed that in the first iteration the best architecture was produced by the application of LDA with a combination of parameters of alpha 0.001, beta 0.001, and number of topics 15. The architecture had 67,5% accuracy level. From the best architecture then training data input is given based on each product review category. The result showed that the combination of the usage of general data and per category data indicate an increase in the average accuracy of 0,82 % from the three-test data. Therefore, in order to produce the best performance of building a classification model of sentiment analysis of the user reviews, it should be performed by applying LDA with a combination of general data and per category data usage
{"title":"Aspect Based Sentiment Analysis in E-Commerce User Reviews Using Latent Dirichlet Allocation (LDA) and Sentiment Lexicon","authors":"Eko Wahyudi, R. Kusumaningrum","doi":"10.1109/ICICoS48119.2019.8982522","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982522","url":null,"abstract":"User ratings on products sold bye-commerce greatly influence the number of purchases. Positive ratings will encourage other buyers to participate in buying the product. While negative ratings given by users will reduce the interest in purchasing products. Nonconformities between rating and user reviews sometimes provide a wrong assessment of a product. This happens because buyers also provide reviews on the quality of delivery services from e-commerce. Based on that issue, the utilization of the Latent Dirichlet Allocation (LDA) could be used on sentiment analysis of the user reviews. Sentiment analysis of the user reviews aims to facilitate e-commerce in informing the product quality as rating supporters that have been given by users. This research aims to determine the classification performance of sentiment analysis on e-commerce user reviews using the LDA algorithm with input data in the form of e-commerce user reviews. Then, compare the application of sentiment analysis of the user reviews with the use of general training data and per category training data. The result of this research showed that in the first iteration the best architecture was produced by the application of LDA with a combination of parameters of alpha 0.001, beta 0.001, and number of topics 15. The architecture had 67,5% accuracy level. From the best architecture then training data input is given based on each product review category. The result showed that the combination of the usage of general data and per category data indicate an increase in the average accuracy of 0,82 % from the three-test data. Therefore, in order to produce the best performance of building a classification model of sentiment analysis of the user reviews, it should be performed by applying LDA with a combination of general data and per category data usage","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"17 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":"124255246","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/ICICoS48119.2019.8982492
Zolanda Anggraeni, H. A. Wibawa
Diabetic retinopathy is a health problem that cause damage to the retinal blood vessels and occurs in more than half of people who suffer from diabetes. It is estimated that around 28 million people experience loss of sight for this reason. Thus, the system for detecting early signs of diabetic retinopathy will be very helpful and one of first signs of the onset of symptoms of diabetic retinopathy is the appearance of exudates in the retinal image of the eye. To build an exudate emergence detection system, in this study use the method of extreme learning machine (ELM) which has a fast learning speed. This system uses the gray level co-occurrence matrix feature extraction with 6 features, namely contrast, homogeneity, correlation, ASM, energy and dissimilarity. To get the best model, six scenarios are used by distinguishing the preprocessing flow. The pre processing stage carried out by all scenarios is optic disc removal, green channel separation, contrast limited adaptive histogram equalization (CLAHE) followed by two different preprocessing lines, namely applying brightness and dilation and erosion operations. Then the second path is radon transform, top-hat filtering, discrete wavelet transform and dilation and erosion. The best model results reached the best accuracy value of 65% with a combination of multiquadric activation functions and 30 hidden neurons.
{"title":"Detection of the Emergence of Exudate on the Image of Retina Using Extreme Learning Machine Method","authors":"Zolanda Anggraeni, H. A. Wibawa","doi":"10.1109/ICICoS48119.2019.8982492","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982492","url":null,"abstract":"Diabetic retinopathy is a health problem that cause damage to the retinal blood vessels and occurs in more than half of people who suffer from diabetes. It is estimated that around 28 million people experience loss of sight for this reason. Thus, the system for detecting early signs of diabetic retinopathy will be very helpful and one of first signs of the onset of symptoms of diabetic retinopathy is the appearance of exudates in the retinal image of the eye. To build an exudate emergence detection system, in this study use the method of extreme learning machine (ELM) which has a fast learning speed. This system uses the gray level co-occurrence matrix feature extraction with 6 features, namely contrast, homogeneity, correlation, ASM, energy and dissimilarity. To get the best model, six scenarios are used by distinguishing the preprocessing flow. The pre processing stage carried out by all scenarios is optic disc removal, green channel separation, contrast limited adaptive histogram equalization (CLAHE) followed by two different preprocessing lines, namely applying brightness and dilation and erosion operations. Then the second path is radon transform, top-hat filtering, discrete wavelet transform and dilation and erosion. The best model results reached the best accuracy value of 65% with a combination of multiquadric activation functions and 30 hidden neurons.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"265 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":"124316218","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/ICICoS48119.2019.8982411
Yasir Abdur Rohman, R. Kusumaningrum
Twitter active users in Indonesia reached 50 million users from a total worldwide of 284 million in 2015. In January 2019, active users on Twitter increased by 52% compared to 2018 where active users were only 27%.A large number of users causes the number of tweet documents increases. Tweet documents that contain information such as user activity, news, story can be processed into valuable information for journalists. All of the information collected then arranged based on related tweets into a storytelling that will become news/article. The whole process is still done manually by collecting one by one for each tweet and most of the tweet documents are collected from the trending topic. Actually, that should be done automatically by collecting tweets that have the same topic. Therefore, this research proposes a method of Twitter storytelling generator that combines Latent Dirichlet Allocation (LDA) and Hidden Markov Model POS-TAG (Part-of-Speech Tagging), so it can generate twitter storytelling based on the certain topic. We implemented two scenarios of the experiment. The first experimental calculating the value of perplexity on LDA and HMM POS-TAG, yielding the lowest perplexity value of 6.31 with alpha 0.001, beta 0.001, and the number of topics 4. While the second experimental calculating the value of ROUGE-1, ROUGE-2, BLEU-1, and BLEU-2 on the results of Twitter storytelling generator, yielding the best ROUGE-1 value is 0.470 with the beta cap value of 0.1 and the best ROUGE-2 value is 0.149 with the beta cap value of 0.001. Meanwhile, the best BLEU-1 value is 0.617 on the topic 1 and the best BLEU-2 value is 0.432 on the topic 3. Twitter storytelling generator using the proposed method has good performance when HMM POS-TAG can tagging the tweet documents correctly.
{"title":"Twitter Storytelling Generator Using Latent Dirichlet Allocation and Hidden Markov Model POS-TAG (Part-of-Speech Tagging)","authors":"Yasir Abdur Rohman, R. Kusumaningrum","doi":"10.1109/ICICoS48119.2019.8982411","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982411","url":null,"abstract":"Twitter active users in Indonesia reached 50 million users from a total worldwide of 284 million in 2015. In January 2019, active users on Twitter increased by 52% compared to 2018 where active users were only 27%.A large number of users causes the number of tweet documents increases. Tweet documents that contain information such as user activity, news, story can be processed into valuable information for journalists. All of the information collected then arranged based on related tweets into a storytelling that will become news/article. The whole process is still done manually by collecting one by one for each tweet and most of the tweet documents are collected from the trending topic. Actually, that should be done automatically by collecting tweets that have the same topic. Therefore, this research proposes a method of Twitter storytelling generator that combines Latent Dirichlet Allocation (LDA) and Hidden Markov Model POS-TAG (Part-of-Speech Tagging), so it can generate twitter storytelling based on the certain topic. We implemented two scenarios of the experiment. The first experimental calculating the value of perplexity on LDA and HMM POS-TAG, yielding the lowest perplexity value of 6.31 with alpha 0.001, beta 0.001, and the number of topics 4. While the second experimental calculating the value of ROUGE-1, ROUGE-2, BLEU-1, and BLEU-2 on the results of Twitter storytelling generator, yielding the best ROUGE-1 value is 0.470 with the beta cap value of 0.1 and the best ROUGE-2 value is 0.149 with the beta cap value of 0.001. Meanwhile, the best BLEU-1 value is 0.617 on the topic 1 and the best BLEU-2 value is 0.432 on the topic 3. Twitter storytelling generator using the proposed method has good performance when HMM POS-TAG can tagging the tweet documents correctly.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"6 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":"124360147","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/ICICoS48119.2019.8982447
Puteri Khatya Fahira, A. Wibisono, H. Wisesa, Zulia Putri Rahmadhani, P. Mursanto, A. Nurhadiyatna
Indonesia is a country rich in culture. One of Indonesia's culturaldiversity is on traditional foods. Traditional food not only has a role in the cultural aspect, but also has an influence on biodiversity. Unfortunately, the current diet of people endangers the existence of traditional foods, which indirectly will also affect Indonesia's food security. Indonesia Local Food Database is one solution proposed to prevent this problem, where the database will play a role to monitor food systems in Indonesia. In this research, database development will focus on collecting data for Sumatra traditionalfood, and also building a model for image classification which will later become one of the main features of the database. Some features like color and texture are extracted from the image. These features are used for classification using 5 classical machine learning models. Evaluation results show performance that as good as deep learning approach.
{"title":"Sumatra Traditional Food Image Classification Using Classical Machine Learning","authors":"Puteri Khatya Fahira, A. Wibisono, H. Wisesa, Zulia Putri Rahmadhani, P. Mursanto, A. Nurhadiyatna","doi":"10.1109/ICICoS48119.2019.8982447","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982447","url":null,"abstract":"Indonesia is a country rich in culture. One of Indonesia's culturaldiversity is on traditional foods. Traditional food not only has a role in the cultural aspect, but also has an influence on biodiversity. Unfortunately, the current diet of people endangers the existence of traditional foods, which indirectly will also affect Indonesia's food security. Indonesia Local Food Database is one solution proposed to prevent this problem, where the database will play a role to monitor food systems in Indonesia. In this research, database development will focus on collecting data for Sumatra traditionalfood, and also building a model for image classification which will later become one of the main features of the database. Some features like color and texture are extracted from the image. These features are used for classification using 5 classical machine learning models. Evaluation results show performance that as good as deep learning approach.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"22 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":"117102385","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}
Finding the right information in easily manner is a need of the community. Virtual assistant is a mobile application to fill this necessity in easily, effective, and efficient manners. Vanika was the virtual assistant developed by Soegijapranata Catholic University (SCU) to serve students and prospective students in finding the right information about the SCU. This study aims to examine the acceptance of virtual assistant-Vanika among them. The answers of two hundred questionnaires were collected from students and prospective students of SCU after they utilized Vanika at least twice. The result reveals that all of factors Information Service Quality, Perceived Ease of Use, and Perceived Usefulness have statistically correlation on Behavioral Intention and each other. Only Information Service Quality has statistically significant direct effect on Behavioral Intention. Prospective students frequently spend more time in searching university information than students, conversely, student perceived easier in using Vanika than prospective students. The last, males spend more time in searching university information than females, and females have more intention to use Vanika than males. The practical implications derived from the findings are the actions to increase individual's acceptance to use Vanika by means of increasing factors relating to Information Service Quality, Perceived Ease of Use, Perceived Usefulness, Age, and Gender.
{"title":"Examining the Acceptance of Virtual Assistant - Vanika for University Students","authors":"Devina Gunadi, Ridwan Sanjaya, Bernardinus Harnadi","doi":"10.1109/ICICoS48119.2019.8982513","DOIUrl":"https://doi.org/10.1109/ICICoS48119.2019.8982513","url":null,"abstract":"Finding the right information in easily manner is a need of the community. Virtual assistant is a mobile application to fill this necessity in easily, effective, and efficient manners. Vanika was the virtual assistant developed by Soegijapranata Catholic University (SCU) to serve students and prospective students in finding the right information about the SCU. This study aims to examine the acceptance of virtual assistant-Vanika among them. The answers of two hundred questionnaires were collected from students and prospective students of SCU after they utilized Vanika at least twice. The result reveals that all of factors Information Service Quality, Perceived Ease of Use, and Perceived Usefulness have statistically correlation on Behavioral Intention and each other. Only Information Service Quality has statistically significant direct effect on Behavioral Intention. Prospective students frequently spend more time in searching university information than students, conversely, student perceived easier in using Vanika than prospective students. The last, males spend more time in searching university information than females, and females have more intention to use Vanika than males. The practical implications derived from the findings are the actions to increase individual's acceptance to use Vanika by means of increasing factors relating to Information Service Quality, Perceived Ease of Use, Perceived Usefulness, Age, and Gender.","PeriodicalId":105407,"journal":{"name":"2019 3rd International Conference on Informatics and Computational Sciences (ICICoS)","volume":"22 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":"125933550","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：