Garuda - Garba Rujukan Digital

Sistem Stemming Otomatis untuk Kata dalam Bahasa Indonesia Rila Mandala; Erry Koryanti; Rinaldi Munir; Harlili Harlili
Seminar Nasional Aplikasi Teknologi Informasi (SNATI) 2004
Publisher : Jurusan Teknik Informatika, Fakultas Teknologi Industri, Universitas Islam Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Stemming is a process to restore words to its base form, by stripping each word fromits derivational and affixes. A stemming process has an important role for machinetranslationand other computational lingustics area. In Malaysian there is a stemmingalgorithm that has been developed and tested for application in information retrieval which isknown as Othman algorithm. There are several differences of Bahasa Indonesia’smorphology and Malay’s morphology, so The Othman algorithm can not be applied directlyin bahasa Indonesia. Furthermore, the accuracy of Othman algorithm also is not good. Thispaper proposes some modifications from Othman algorithm. The modifications includes,various stemming procedures, rule of affixes, and dictionary of root words. Experiments showthat Our modification method has a better accuracy in stemming Bahasa Indonesia’s words.Keywords: stemming, word-lemmatization, affix-stripping

ANT-WUM: ALGORITMA BERBASIS ANT COLONY OPTIMIZATION UNTUK WEB USAGE MINING Abdurrahman -; Bambang Riyanto Trilaksono; Rila Mandala; Rajesri Govindaraju
JURNAL TEKNOLOGI TECHNOSCIENTIA Technoscientia Vol 2 No 1 Agustus 2009
Publisher : Lembaga Penelitian & Pengabdian Kepada Masyarakat (LPPM), IST AKPRIND Yogyakarta

This paper is continuity research from our previous work in Ant-Miner implementation for web user classification. In our previous work, we implemented Ant-Miner algorithm for web user classification same with Ant-Miner for classification task in data mining domain. In this paper, we propose modification of heuristic function of Ant-Miner based on web usage mining (WUM) problem, that we name Ant-WUM. The heuristic function ACO is based on local problem domain. Information theory is common heuristic function used in classification task, such as implemented in C4.5 algorithm and ant-miner algorithm. Ant-WUM uses heuristic function based on closeness principle that implemented in clustering problem in WUM. We propose to use data from web access log, profile user, and transaction data to provide some attributes as term candidate of classification rule by Ant-WUM algorithm. We compared Ant-WUM algorithm with Ant-Miner algorithm. The result indicates that Ant-WUM has competitive result in term of accuracy rate, amount of rules, and computation time.

Robust Automatic Speech Recognition Features using Complex Wavelet Packet Transform Coefficients Tjong Wan Sen; Bambang Riyanto Trilaksono; Arry Akhmad Arman; Rila Mandala
Journal of ICT Research and Applications Vol. 3 No. 2 (2009)
Publisher : LPPM ITB

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.5614/itbj.ict.2009.3.2.4

To improve the performance of phoneme based Automatic Speech Recognition (ASR) in noisy environment; we developed a new technique that could add robustness to clean phonemes features. These robust features are obtained from Complex Wavelet Packet Transform (CWPT) coefficients. Since the CWPT coefficients represent all different frequency bands of the input signal, decomposing the input signal into complete CWPT tree would also cover all frequencies involved in recognition process. For time overlapping signals with different frequency contents, e. g. phoneme signal with noises, its CWPT coefficients are the combination of CWPT coefficients of phoneme signal and CWPT coefficients of noises. The CWPT coefficients of phonemes signal would be changed according to frequency components contained in noises. Since the numbers of phonemes in every language are relatively small (limited) and already well known, one could easily derive principal component vectors from clean training dataset using Principal Component Analysis (PCA). These principal component vectors could be used then to add robustness and minimize noises effects in testing phase. Simulation results, using Alpha Numeric 4 (AN4) from Carnegie Mellon University and NOISEX-92 examples from Rice University, showed that this new technique could be used as features extractor that improves the robustness of phoneme based ASR systems in various adverse noisy conditions and still preserves the performance in clean environments.

ANT-WUM: ALGORITMA BERBASIS ANT COLONY OPTIMIZATION UNTUK WEB USAGE MINING Abdurrahman -; Bambang Riyanto Trilaksono; Rila Mandala; Rajesri Govindaraju
JURNAL TEKNOLOGI TECHNOSCIENTIA Technoscientia Vol 2 No 1 Agustus 2009
Publisher : Lembaga Penelitian & Pengabdian Kepada Masyarakat (LPPM), IST AKPRIND Yogyakarta

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34151/technoscientia.v2i1.67

This paper is continuity research from our previous work in Ant-Miner implementation for web user classification. In our previous work, we implemented Ant-Miner algorithm for web user classification same with Ant-Miner for classification task in data mining domain. In this paper, we propose modification of heuristic function of Ant-Miner based on web usage mining (WUM) problem, that we name Ant-WUM. The heuristic function ACO is based on local problem domain. Information theory is common heuristic function used in classification task, such as implemented in C4.5 algorithm and ant-miner algorithm. Ant-WUM uses heuristic function based on closeness principle that implemented in clustering problem in WUM. We propose to use data from web access log, profile user, and transaction data to provide some attributes as term candidate of classification rule by Ant-WUM algorithm. We compared Ant-WUM algorithm with Ant-Miner algorithm. The result indicates that Ant-WUM has competitive result in term of accuracy rate, amount of rules, and computation time.

Comparison of Adaptive Boosting and Bootstrap Aggregating Performance to Improve the Prediction of Bank Telemarketing Agus Priyanto; Rila Mandala
IT for Society Vol 5, No 2 (2020)
Publisher : President University

Background: Telemarketing is an effectivemarketing strategy lately, because it allows long-distanceinteraction making it easier for marketing promotionmanagement to market their products. But sometimes withincessant phone calls to clients that are less potential to causeinconvenience, so we need predictions that produce goodprobabilities so that it can be the basis for making decisionsabout how many potential clients can be contacted whichresults in time and costs can be minimized, telephone calls canbe more effective, client stress and intrusion will be reduced.strong.Method: This study will compare the classificationperformance of Bank Marketing datasets from the UCIMachine Learning Repository using data mining with theAdaboost and Bagging ensemble approach, base algorithmusing J48 Weka, and Wrapper subset evaluation featureselection techniques and previously data balancing wasperformed on the dataset, where the expected results can beknown the best ensemble method that produces the bestperformance of both.Results: In the Bagging experiment, the best performanceof Adaboost and J48 with an accuracy rate of 86.6%, Adaboost83.5% and J48 of 85.9%Conclusion: The conclusion obtainedfrom this study that the use of data balancing and featureselection techniques can help improve classificationperformance, Bagging is the best ensemble algorithm from thisstudy, while for Adaboost is not productive for this studybecause the basic algorithm used is a strong learner whereAdaboost has Weaknesses to improve strong basic algorithm.

IMPLEMENTATION OF K-MEANS ALGORITHM FOR INFORMATION TECHNOLOGY FRESHMAN CLASS DIVISION Arfan As’Sidiq; Rila Mandala
IT for Society Vol 4, No 1 (2019)
Publisher : President University

Almost all universities divide their IT freshman into classes randomly or based on students score, either their score during the selection test held by the university or National Examination score. Universities often find case that a class consists of all ‘smart’ students and a class consists of all ‘lazy’ students. This thesis intends to create an application to help universities divides their Information Technology freshman into classes based on freshman competency and experience about Information Technology (IT) on the senior high school. The experiment is conducted by collecting data IT students who are not in the first semester. The data consists of their experience about IT as well as other knowledge fields and their current GPA. The results of the experiment show that from 50 data samples collected, the application correctly predicts 34 students GPA range based on respondents competency with IT and other knowledge fields during their study in senior high school.

Prediksi Harga Minyak Kelapa Sawit Dalam Investasi Dengan Membandingkan Algoritma Naïve Bayes, Support Vector Machine dan K-Nearest Neighbor Deny Haryadi; Rila Mandala
IT for Society Vol 4, No 1 (2019)
Publisher : President University

Harga minyak kelapa sawit bisa mengalami kenaikan, penurunan maupun tetap setiap hari karena faktor yang mempengaruhi harga minyak kelapa sawit seperti harga minyak nabati lain (minyak kedelai dan minyak canola), harga minyak mentah dunia, maupun nilai tukar riil antara kurs dolar terhadap mata uang negara produsen (rupiah, ringgit, dan canada) atau mata uang negara konsumen (rupee). Untuk itu dibutuhkan prediksi harga minyak kelapa sawit yang cukup akurat agar para investor bisa mendapatkan keuntungan sesuai perencanaan yang dibuat. tujuan dari penelitian ini yaitu untuk mengetahui perbandingan accuracy, precision, dan recall yang dihasilkan oleh algoritma Naïve Bayes, Support Vector Machine, dan K-Nearest Neighbor dalam menyelesaikan masalah prediksi harga minyak kelapa sawit dalam investasi. Berdasarkan hasil pengujian dalam penelitian yang telah dilakukan, algoritma Support Vector Machine memiliki accuracy, precision, dan recall dengan jumlah paling tinggi dibandingkan dengan algoritma Naïve Bayes dan algoritma K-Nearest Neighbor. Nilai accuracy tertinggi pada penelitian ini yaitu 82,46% dengan precision tertinggi yaitu 86% dan recall tertinggi yaitu 89,06%.

Analisa Judul Skripsi untuk Menentukan Peminatan Mahasiswa Menggunakan Vector Space Model dan Metode K-Nearest Neighbor Dewi Marini Umi Atmaja; Rila Mandala
IT for Society Vol 4, No 2 (2019)
Publisher : President University

Sulitnya menentukan klasifikasi judul skrpsi berdasarkan peminatan yang diambil oleh mahasiswa informatika unjani merupakan salah satu permasalahan penting yang dihadapi oleh pihak Jurusan. Tujuan dari penelitian ini yaitu memberikan sebuah penunjang keputusan bagi pihak Jurusan agar setiap judul skripsi yang diajukan oleh mahasiswa sesuai dengan peminatan. Berdasarkan hasil penelitian yang telah dilakukan, model yang dibangun menggunakan algoritma KNN menghasilkan tingkat akurasi yang lebih tinggi jika dibandingkan dengan model yang dibangun menggunakan algoritma VSM. Nilai akurasi tertinggi berdasarkan hasil pengujian pada penelitian ini adalah sebasar 96,85%.

COMPARISON INTENT RECOGNITION ON FOOD DELIVERY SERVICE COMPLAINT IN TWITTER WITH RECURRENT AND CONVOLUTIONAL NEURAL NETWORK Irfan Nasrullah; Rila Mandala
IT for Society Vol 5, No 1 (2020)
Publisher : President University

In this research, the case of intent classification for Customer Relation Management (CRM) how to handle complaints as a domain to be followed up, where datasets are extracted from the conversation on Twitter. The research objectives support three key findings to comparing the CNNs and BRNNs model to intent recognition by vectorization text: (1) Which architecture performs better (accuracy) depends on how important it is to semantically understand the whole sequence and (2) Learning rate changes performance relatively smoothly, while the optimal result iterated by change hidden size and batch size result in large fluctuations. (3) Last, how word vectorization is able to define sub-domain of the complaints by word vector classification.

Peningkatan Sarana Pembelajaran Daring Dengan Optimalisasi Penggunaan Video Editing Di SMPIT Annida Rosalina Rosalina; Genta Sahuri; Chong Min An; Rila Mandala
Jurnal Pengabdian Masyarakat Nusantara Vol. 3 No. 1 (2023): Februari-Juli 2023
Publisher : Lembaga Otonom Lembaga Informasi dan Riset Indonesia (KITA INFO dan RISET)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35870/jpmn.v3i1.846

The advantage of video media in education is that it can boost students' interest in the subject matter by allowing them to watch and hear. Teachers are expected to be able to develop engaging learning materials that encourage students to participate actively in their education. This PkM activity was completed at SMPIT Annida as an activity partner. The issues that were identified were 1) The teacher's ability to self-produce learning media, particularly in making learning videos, 2) The availability of free software that could be used to produce video learning media, but never utilized by teachers, and 3) the improvement process for teaching and learning that is effective, efficient, and more engaging. The focus of this activity was on the instructors and students.

Title

Found 31 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 31 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 31 Documents
Search