IJCCS (Indonesian Journal of Computing and Cybernetics Systems)
Not a member yet
480 research outputs found
Sort by
Indonesian Music Classification on Folk and Dangdut Genre Based on Rolloff Spectral Feature Using Support Vector Machine (SVM) Algorithm
Get PDFMusic Genre Classification is one of the interesting digital music processing topics. Genre is a category of artistry, in this case, especially music, to characterize and categorize music is now available in various forms and sources. One of the applications is in determining the music genre classification on folk songs and dangdut songs.The main problem in the classification music genre is to find a combination of features and classifiers that can provide the best result in classifying music files into music genres. So we need to develop methods and algorithms that can classify genres appropriately. This problem can be solved by using the Support Vector Machine (SVM). The genre classification process begins by selecting the song file that will be classified by the genre, then the preprocessing process, the collection features by utilizing feature extraction, and the last process is Support Vector Machine (SVM) classification process to produce genre types from selected song files. The final result of this research is to classify Indonesian folk music genre and dangdut music genre along with the 83.3% accuracy values that indicate the level of system relevance to the results of music genre classification and to provide genre labels on music files as to facilitate the management and search of music files
Sunspot Number Prediction Using Gated Recurrent Unit (GRU) Algorithm
Get PDFSunspot is an area on photosphere layer which is dark-colored. Sunspot is very important to be researched because sunspot is affected by sunspot numbers, which present the level of solar activity. This research was conducted to make prediction on sunspot numbers using Gated Recurrent Unit (GRU) algorithm. The work principle of GRU is similar to Long short-term Memory (LSTM) method: the information from the previous memory is processed through two gates, that is update gate and reset gate, then the output generated will be input for the next process. The purpose of predicting sunspot numbers was to find out the information of sunspot numbers in the future, so that if there is a significant increase in sunspot numbers, it can inform other physical consequences that may be caused. The data used was the data of monthly sunspot numbers obtained from SILSO website. The data division and parameters used were based on the results of the trials resulted in the smallest MAPE value. The smallest MAPE value obtained from the prediction was 7.171% with 70% training data, 30% testing data, 150 hidden layer, 32 batch size, 100 learning rate drop.
Hate Speech Detection in Indonesian Twitter using Contextual Embedding Approach
Get PDFHate speech develops along with the rapid development of social media. Hate speech is often issued due to a lack of public awareness of the difference between criticism and statements that might contribute to this crime. Therefore, it is very important to do early detection of sentences that will be written before causing a criminal act due to public ignorance. In this paper, we use the advancement of deep neural networks to predict whether a sentence contains a hate speech and abusive tone. We demonstrate the robustness of different word and contextual embedding to represent the semantic of hate speech words. In addition, we use a document embedding representation via a recurrent neural networks with gated recurrent unit as the main architecture to provide richer representation. Compared to syntactic representation of the previous approach, the contextual embedding in our model proved to give a significant boost on the performance by a significant margin
Decision Support System for Laptop Selection Using AHP Method and Profile Matching
Get PDF Laptop is a desktop personal computer (PC) whose dimensions are reduced to increase flexibility in its use. However, the large number of products will make it difficult for consumers to choose a laptop that suits the needs of consumers who want to buy it.The purpose of this research is to help buyers who want to buy laptop products according to their needs by making a Decision Support System (DSS). There are 12 criteria considered in this research, price, processor, RAM capacity, hard disk capacity, SSD capacity, V-RAM capacity, maximum RAM upgrade capacity, laptop weight, screen size, screen type, screen refresh rate, and screen resolution. Choosing a laptop product there is a criterion value of a laptop product and a value of preference criteria from the buyer as a decision maker. Also the criteria values on laptop products have different contributions to the overall value of the laptop product. Thus, the methods used are Analytical Hierarchy Process (AHP), Profile Matching (PM) with linear interpolation, and Simple Addictive Weighting (SAW) to determine the recommended options. Lastly, SPK that has been made will be able to provide recommendations best alternative choices and best suit the needs of buyers for selecting laptop products
Rule of Land Potential for Paddy Use Rough Set Method
Get PDFBlitar district has become one of the many cities in Java the land situation is largely a good soil of vuikanik to be used as farmland. Agriculture is one of the priority sectors in Blitar district and is supported by culture, geographical conditions and the number of people whose livelihoods are farmers.Hence, it requires a way of knowing where a region might have a potential paddy commodity. It is hoped that the government of blitar will be able to make the best use of the number of paddy commodities produced in blitar district with the many farmers available. A rough set is able to produce information with a rule pattern (rule) which can determine the potential areas for paddy commodities in Blitar district by using factors of harvested area, production amount, and number of farmers per sub-district. This research is not only done analytically but also help from Rosetta's software to test analytic data analysis use rough set. The result of this study is rule as many as 38 rule that can explain the possibility of stake based on the 3 decision attributes: potential, low potential, and not potential. For those areas there is a good chance paddy commodity potential area based on the rules that have been formed is area have a large crop, a large amount of paddy produced, and a small number of farmers
Detection of Cataract Based on Image Features Using Convolutional Neural Networks
Get PDFCataract are the highest cause of blindness that there are 32.4 million people experiencing blindness and as many as 191 million people experiencing visual disabilities in 2010 in the world. On the other hand, the longer a patient suffers from cataracts or late treatment. The development of cataract identification using a traditional algorithm based on feature representation is highly dependent on the classification process carried out by an eye specialist so that the method is prone to misclassification of a person detected or not. However, at this time there is a deep learning, convolutional neural network (CNN) which is used for pattern recognition which can help automate image classification. This research was conducted to increase the accuracy value and minimize data loss in the process of cataract identification by performing an experience namely the manipulation process was carried out by changing epochs. The results of this study indicate that the addition of epochs affects accuracy and loss data from CNN. By comparing variety of epoch values it can be ignored that the higher the age values used, the higher the value of the model. In this study, using the epoch 50 value reached the highest value with a value of 95%. Based on the model that has been made it has also been successful to receive images according to the specified class. After testing accurately, 10 images achieved an average accuracy of 88%
Decision Support System For Determining Campus Promotion Media In New Student Admissions With Analytical Network Process And Regression Methods
Get PDFCampus promotion is carried out in an effort to introduce the campus to the community, especially prospective new students. This effort was carried out as an action that was considered effective in recruiting new students. Various obstacles experienced by tertiary institutions in implementing campus promotion, namely the lack of need for supporting funds, limited human resources (HR), the right decision system for the selection of promotional media. This study analyzes the decision support system in selecting the right promotional media for campus promotion. The research objective is to assist campus management in selecting the right promotional media with a decision support system for determining the promotion media for new student admissions and determining the priority of the promotional media that will be used by private universities (PTS) in the city of Kendari. The sample in this study amounted to 40 respondents from 24 universities. The method used is the Analytical Network Process (ANP) method and the Regression method uses factor analysis. The results of the research analysis show that promotional media using the campus website has a number one rating, namely with a value of 26.2% while word of mouth has a second rating of 23.3%, then social media with a score of 23.1%, brochure 8.9%, print media and electronic media with a value of 6.2% and billboards have a value of 5.7%
Online Learning Video Recommendation System Based on Course and Sylabus Using Content-Based Filtering
Get PDFLearning using video media such as watching videos on YouTube is an alternative method of learning that is often used. However, there are so many learning videos available that finding videos with the right content is difficult and time-consuming. Therefore, this study builds a recommendation system that can recommend videos based on courses and syllabus. The recommendation system works by looking for similarity between courses and syllabus with video annotations using the cosine similarity method. The video annotation is the title and description of the video captured in real-time from YouTube using the YouTube API. This recommendation system will produce recommendations in the form of five videos based on the selected courses and syllabus. The test results show that the average performance percentage is 81.13% in achieving the recommendation system goals, namely relevance, novelty, serendipity and increasing recommendation diversity
Transfer Learning of Pre-trained Transformers for Covid-19 Hoax Detection in Indonesian Language
Get PDFNowadays, internet has become the most popular source of news. However, the validity of the online news articles is difficult to assess, whether it is a fact or a hoax. Hoaxes related to Covid-19 brought a problematic effect to human life. An accurate hoax detection system is important to filter abundant information on the internet. In this research, a Covid-19 hoax detection system was proposed by transfer learning of pre-trained transformer models. Fine-tuned original pre-trained BERT, multilingual pre-trained mBERT, and monolingual pre-trained IndoBERT were used to solve the classification task in the hoax detection system. Based on the experimental results, fine-tuned IndoBERT models trained on monolingual Indonesian corpus outperform fine-tuned original and multilingual BERT with uncased versions. However, the fine-tuned mBERT cased model trained on a larger corpus achieved the best performance
Covid-19 Hoax Detection Using KNN in Jaccard Space
Get PDFSocial media has become a communication key to spark thinking, dialogue and action around social issues. Hoax is information that added or subtracted from the content of the actual news. The spread of unconfirmed Covid-19 news can cause public concern. The purpose of this research was to modify KNN with Jaccard Space in the classification of hoax news related to Covid-19. The data used from Jabar Saber Hoaks and Jala Hoaks. The classification results with KNN with Jaccard Space and stemming Nazief & Adriani get the highest accuracy than other models in this research. The accuracy of the KNN model on the Jaccard Space with stemming Nazief & Adriani and K = 5 was 75.89%, while for Naïve Bayes was 65.18%