JUTI: Jurnal Ilmiah Teknologi Informasi
Not a member yet
407 research outputs found
Sort by
IMPROVED LIP-READING LANGUAGE USING GATED RECURRENT UNITS
Full text linkLip-reading is one of the most challenging studies in computer vision. This is because lip-reading requires a large amount of training data, high computation time and power, and word length variation. Currently, the previous methods, such as Mel Frequency Cepstrum Coefficients (MFCC) with Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN) with LSTM, still obtain low accuracy or long-time consumption because they use LSTM. In this study, we solve this problem using a novel approach with high accuracy and low time consumption. In particular, we propose to develop lip language reading by utilizing face detection, lip detection, filtering the amount of data to avoid overfitting due to data imbalance, image extraction based on CNN, voice extraction based on MFCC, and training model using LSTM and Gated Recurrent Units (GRU). Experiments on the Lip Reading Sentences dataset show that our proposed framework obtained higher accuracy when the input array dimension is deep and lower time consumption compared to the state-of-the-art
PREDICTION OF MULTIVARIATE TIME SERIES DATA USING ECHO STATE NETWORK AND HARMONY SEARCH
Full text linkMultivariate time series data prediction is widely applied in various fields such as industry, health, and economics. Several methods can form prediction models, such as Artificial Neural Network (ANN) and Recurrent Neural Network (RNN). However, this method has an error value more significant than the development method of RNN, namely the Echo State Network (ESN). The ESN method has several global parameters, such as the number of reservoirs and the leaking rate. The determination of parameter values dramatically affects the performance of the resulting prediction model. The Harmony Search (HS) optimization method is proposed to provide a solution for determining the parameters of the ESN method. The HS method was chosen because it is easier to implement, and based on other research, the HS method gets the optimum value better than other meta-heuristic methods. The methods compared in this study are RNN, ESN, and ESN-HS. Root Mean Square Error (RMSE) and Mean Absolute Percent Error (MAPE) are used to measure the error rate of forecasting results. ESN got a smaller error value than RNN, and ESN-HS produced a minor error value among the other trials, namely 0.782e-5 for RMSE and 0.28% for MAPE. The HS optimization method has successfully obtained the appropriate global parameters for the ESN prediction model
SELENIUM FRAMEWORK FOR WEB AUTOMATION TESTING: A SYSTEMATIC LITERATURE REVIEW
Full text linkSoftware Testing plays a crucial role in making high-quality products. The process of manual testing is often inaccurate, unreliable, and needed more than automation testing. One of these tools, Selenium, is an open-source framework that used along with different programming languages: (python, ruby, java, PHP, c#, etc.) to automate the test cases of web applications. The purpose of this study is to summarize the research in the area of selenium automation testing to benefit the readers in designing and delivering automated software testing with Selenium. We conducted the standard systematic literature review method employing a manual search of 2408 papers, and applying a set of inclusion/exclusion criteria the final literature included 16 papers published between 2009 and 2020. The result is using Selenium as a UI for web automation, not only all of the app functionality that has been tested, But also it can be applied with added some method or other algorithms like data mining, artificial intelligence, and machine learning. Furthermore, it can be implemented for security testing. In the future research for selenium framework automation testing, the implementation should more focus on finding effective and maintainability on the application of Selenium in other methodologies and is applied with the better improvement that can be matched for web automation testing
EMERGENCY PROCESSES HANDLING IN URBAN AREA USING MODIFIED DIJKSTRA METHOD
Full text linkEmergency Aid has a very vital role in saving the patient\u27s life. The emergency process involves two stages, namely the pre-hospital and the hospital stage. Striving for the entire emergency process is to have the fastest response time. The initial part of emergency treatment (pre-hospital) is determining the shortest and fastest route to the hospital. In addition, the availability of the targeted hospital must also be considered. We modified Dijkstra\u27s Algorithm to produce the shortest route and the fastest time by considering the availability of the targeted hospital to support the handling of the emergency process. The modification made to the Dijkstra algorithm replaces the weight of Dijkstra\u27s distance with a quantity representing the congestion rate and distance. Besides, the event time is estimated to determine the status of the intended hospital. As a result, Dijkstra\u27s modification method can produce a more efficient and faster route
AUTOMATIC TESTING FRAMEWORK BASED ON SERENITY AND JENKINS AUTOMATED BUILD
Full text linkSoftware Testing plays an important role in making high quality products and the right time. The process of testing done manually is often inaccurate, unreliable, and needed more than automatic testing. This research proposes a new framework for automation testing. This framework will help developers to create applications with better quality and shorten testing time. This framework offers a solution for developers so that the testing process is carried out easily and quickly. Our proposed concept consists of an automated test script based on Serenity Framework and can be done as a background process using Jenkins. Input of the system is a testing scenario, then mapped into Java Programming Language. Output of this system are test reports that represent the scenario that has been carried out. the results of implementation system prove that developers are helped by this framework in the software testing process. So that in this study it can be concluded that the automated testing framework that has been developed can improve the quality of application products through effective and efficient work methods
EFFICIENCY OF FLOODING BY DEVELOPING RELIABLE SUBNETWORK METHODS ON FIBBING ARCHITECTURE IN THE HYBRID ENVIRONMENT SDN
Full text linkIn the technology world especially in the field of current network of Autonomous Systems connectivity (AS) is indispensable. Especially against the dynamic routing protocols that are often used compared to static routing protocols. In supporting this current network, it takes efficient and effective routing protocols capable of covering a sizable scale. Software Defined Network (SDN) is a technological innovation in the network world that has a separate Control Plane and Data Plane that makes it easy to configure on the Control Plane side. Control Plane is the focal point on a process of bottleneck in SDN architecture. Performance is a critical issue in large-scale network implementations because of the large demand load occurring in the Control Plane by generating low throughput value. This research will be conducted testing on the Hybrid network of SDN by using OSPF routing protocol, based on the Fibbing architecture implemented on the system network Hybrid SDN also able to assist in improving performance, but there are constraints when sending flooding which is used as a fake node forming. Many nodes are not skipped as distribution lines in the formation of a fake node, in which case it will certainly affect the value of throughput to be unstable and decrease. This can be overcome by using the Isolation Domain method to manage the LSA Type-5 flooding efficiency
IMPLEMENTATION OF MULTILAYER PERCEPTRON FOR STUDENT FAILURE PREDICTION
Full text linkUniversity is one of the educational institutions and can be established by the government or the individual. At this time, Indonesia has hundreds of universities spread throughout the region. As an educational institution, university of course must be able to educate its students and issue quality graduates with the academically and non-academically qualified. In its implementation, there are many problems that should be resolved as well as possible, such as when there are students who intentionally stop or disappear before completing their education or are even unable to complete their education and issued by institution (dropout).Based on these problems, this research makes a model for predicting students who have the potential to fail or dropout during their studies using one of the data mining methods namely Multilayer Perceptron by referring to personal and academic data. The results obtained from this research are 86.9% an accuracy rate with the 54.7% sensitivity, and 95.4% specificity. This research is expected to be used to determine the need strategies to minimize the number of students who stop or dropout
CONTINUOUS MULTIQUERIES K-DOMINANT SKYLINE ON ROAD NETWORK
Full text linkThe increasing use of mobile devices makes spatial data worthy of consideration. To get maximum results, users often look for the best from a collection of objects. Among the algorithms that can be used is the skyline query. The algorithm looks for all objects that are not dominated by other objects in all of its attributes. However, data that has many attributes makes the query output a lot of objects so it is less useful for the user. k-dominant skyline queries can be a solution to reduce the output. Among the challenges is the use of skyline queries with spatial data and the many user preferences in finding the best object. This study proposes IKSR: the k-dominant skyline query algorithm that works in a road network environment and can process many queries that have the same subspace in one processing. This algorithm combines queries that operate on the same subspace and set of objects with different k values by computing from the smallest to the largest k. Optimization occurs when some data for larger k are precomputed when calculating the result for the smallest k so the Voronoi cell computing is not repeated. Testing is done by comparing with the naïve algorithm without precomputation. IKSR algorithm can speed up computing time two to three times compared to naïve algorithm
A MODEL AND IMPLEMENTATION OF ACADEMIC DATA INTEGRATION IN NEAR-REAL TIME USING MESSAGE-ORIENTED MIDDLEWARE TO SUPPORT ANALYSIS OF STUDENT PERFORMANCE IN THE INFORMATION TECHNOLOGY DEPARTMENT OF POLITEKNIK CALTEX RIAU
Full text linkData utilization has effectively contributed for institutions growth by providing insights for managerial purposes. In Information Technology (IT) Department of Politeknik Caltex Riau, information systems were built separately, makes it hard for the head of study program to analyze academic performance. For analytical purposes, there’s a business intelligence developed to equip each head of study programs in IT Department with knowledge about their department. Unfortunately, the business intelligence hasn’t considered with data integration. To solve this problem, this research proposes 2 different academic near-real time data integration model that are documented using Enterprise Integration Pattern and benchmarkes the implementation to obtain best data integration model. The models use Message-Oriented Middleware, a technology that enables asynchronous communication between diverse applications. This research uses WSO2 ESB as the MOM tools in Service-Oriented Architecture (SOA) that use NuSOAP library for helping generating web service WSDL and will use Enterprise Application Integration approach. The testing is conducted based on ISO 9126 aspects: functionality, efficiency, and reliability. Based on the testing results, it can be concluded that both integration models fulfill the functionality and reliability aspects, but the 2nd pattern is more efficient because it distincts message channel and store for each dimension and fact table
PENGARUH INTERFERENSI HIDDEN NODE TERHADAP MODEL PREDIKSI KETERSEDIAAN BANDWIDTH DI JARINGAN NIRKABEL
Full text linkKetersediaan bandwidth merupakan salah satu aspek penting untuk menjamin QoS dalam transmisi data, terutama pada jaringan nirkabel. Walaupun demikian, prediksi ketersediaan pada jaringan nirkabel masih sulit dilakukan karena medium transmisi dapat digunakan oleh beberapa node secara bersamaan. Selain itu jaringan nirkabel juga rentan terhadap pengaruh dari sinyal transmisi yang dihasilkan dari node lain, terutama hidden node. Beberapa penelitian telah dilakukan untuk mengembangkan model prediksi ketersediaan bandwidth. Walaupun demikian, belum terdapat mekanisme terstandarisasi yang digunakan untuk mengevaluasi ketersediaan bandwidth pada jaringan nirkabel. Selain itu tingkat akurasi dari setiap model juga masih belum diketahui ketika diimplementasikan pada jaringan nirkabel, terutama dengan keberadaan hidden node. Oleh karena itu penelitian ini berupaya untuk menginvestigasi kinerja dari setiap model untuk memprediksi ketersediaan bandwidth pada jaringan nirkabel dengan interferensi hidden node. Model prediksi yang dibandingkan adalah Distributed Lagrange Interpolation Based Available Bandwidth Estimation (DLI-ABE), Cognitive Passive Estimation Of The Available Bandwidth (cPEAB), Improved Available Bandwidth (IAB), dan Available Bandwidth Estimation (ABE). Percobaan dilakukan dalam skala simulasi yang dikembangkan menggunakan simulasi jaringan OMNet++. Hasil penelitian menunjukan bahwa model ABE memperoleh tingkat akurasi yang paling baik sebesar 85,25%