No abstract

No abstract

Abstract-Determining the right class on student achievement is important in an evaluation process, because placing students in the right class helps lecturer in reflecting the successfullness of learning process. This problem relates to the performance of classification method which is measured by the classifier metrics. High performance is indicated by the optimality of these classifier's metrics. Besides, meta-algorithm adaptive boosting has been proven to be able to improve the performance of classifier in various fields. Therefore, this paper employs adaptive boosting to reduce the number of incorrect student placement in a class. The experimental results of implementing adaptive boosting in classifying student achievement shows that there is an increase of performance of all classification metrics, i.e., Kappa, Precision, Recall, F-Measure, ROC, and MAE. In terms of accuracy, J-48 is able to rise about 3.09%, which means this method reduces three misclassified students. Additionally, decisionStump increases 12.37% of accuracy. This also means this method is able to decrease 12 misclassified students. Finally, Simple Cart reaches the highest accuracy of about 23.71%, while the number of misclassified students is reduced to 24 students. However, there is no improvement in Random Forest method by using this adaptive boosting. Intisari-Penetapan kelas yang tepat pada prestasi belajar mahasiswa adalah suatu hal yang penting dalam proses evaluasi.Hal ini disebabkan keberadaan mahasiswa pada kelas yang tepat dapat membantu dosen untuk merefleksikan berhasil tidaknya proses belajar mengajar yang telah dilakukan. Masalah ketepatan kelas ini berkaitan dengan kinerja metode klasifikasi yang diukur dengan ukuran-ukuran classifier. Kinerja yang tinggi diindikasikan dengan optimalnya ukuran-ukuran classifier tersebut. Di sisi lain, meta-algoritme adaptive boosting telah terbukti mampu meningkatkan kinerja classifier pada berbagai bidang. Oleh karena itu, makalah ini menerapkan adaptive boosting untuk meningkatkan kinerja classifier agar keberadaan mahasiswa pada kelas yang tidak sesuai bisa direduksi. Hasil uji coba penerapan adaptive boosting pada klasifikasi prestasi belajar mahasiswa ini menunjukkan peningkatan kinerja pada semua ukuran classifier, Penggunaan sistem e-learning dan sistem evaluasi ini akan menghasilkan tumpukan data pada tempat penyimpanan. Tumpukan data ini dapat digunakan untuk pengoptimalan pada proses-proses tersebut dengan penerapan data mining. Hal ini dapat dilakukan karena penerapan ini menghasilkan pengetahuan yang sangat berguna. Pengetahuan ini diperoleh dari beberapa aktivitas dari data mining, antara lain klasifikasi, klaster, dan analisis asosiasi. Beberapa penelitian tentang klasifikasi antara lain prediksi mahasiswa nonaktif dengan menggunakan beberapa metode klasifikasi [6], klasifikasi untuk memprediksi prestasi belajar siswa dengan menggunakan Naïve Bayesian [7], dan klasifikasi dengan decision tree untuk mengetahui hubungan kinerja akademik dan bidang minat [8]. Untuk penelitian pen...

The user's feedback on healthcare services is usually based on ratings from post-service questionnaires. However, in order to get a clear view of the user's perspective, online text reviews need to be analyzed. We combined targeted and aspect-based sentiment analysis by multi-level attention to get a specific user sentiment on a target of an aspect. The multi-level attention consists of Target-level and Sentence-level attention. Our proposed framework is based on Bidirectional Gated Recurrent Unit. Bi-GRU is commonly known to have comparable results compared to LSTM while having lesser computational complexity. We also utilized Bidirectional LSTM based Character-Enhanced Token-Embedding to handle out of vocabulary words and misspelling to avoid error in detecting sentiment. We created a dataset of online healthcare reviews from 2018-2020, targeting the name of the hospital or department, with ten aspects: cleanliness, cost, doctor, food, nurse, parking, receptionist and billing, safety, test and examination, and waiting time. To improve the results of our proposed method, we calculated polarity weight to handle imbalanced aspects in the dataset. We classified these reviews into three polarities, which are positive, negative, and neutral. Based on our experiments, we achieved the best F1-Score of 88%.

No abstract

No abstract

No abstract

No abstract