Hasil Pencarian  ::  Simpan CSV :: Kembali

Abstrak :
Departemen Teknik Elektro sebelumnya sudah mengembangkan sistem penilai esai otomatis (SIMPLE-O) yang berbasis algoritma winnowing dan diterapkan pada bahasa Jepang. Sistem penilai esai otomatis tersebut menggunakan algortima winnowing yang berbasiskan fingerprint dan hashing untuk mendeteksi tingkat kemiripan teks. Sistem tersebut memiliki rata-rata akurasi nilai total seluruh data hingga 90.92% dengan akurasi nilai total perpeserta ujian dapat mencapai 99.91% dan akurasi perjawaban untuk tiap peserta ujian berkisar dari 60.19% hingga 100%. Penelitian kali ini berusaha untuk mencoba untuk menaikkan akurasi tersebut. Cara yang digunakan ialah menganti hashing yang digunakan dari Rolling Hash ke MD5 dan mengimplementasi synonym recognition. Hasil percobaan ini memiliki rata-rata tingkat akurasi 85.61% dengan akurasi perjawaban untuk tiap perserta ujian berkisar 68.44% hingga 99.96%

---

Departement of Electrical Engineering has already developed automatic essay grading system (SIMPLE-O) which utilize winnowing algorithm which is a fingerprint-based and hash-based algorithm for detecting similarity between texts. The system have result of average of total score for all students is 90.92% with accuracy for each student is up to 99.91% and accuracy for each problem ranged from 60.19% to 100%. This research will try to raise the accuracy. The proposed method is by changing the hashing used by the system from Rolling Hash to MD5 and implementing synonym recognition. The result of conducted experiment has the average of accuracy of 85.61% and the accuracy for each problem ranged from 68.44% to 99.96%.

Abstrak :
Simple-O merupakan sebuah sistem penilaian esai otomatis (essay grading) yang meggunakan metode Latent Semantic Analysis (LSA). Prinsip kerja sistem ini adalah dengan melakukan konversi jawaban ke dalam bentuk matriks yang kemudian secara statistik dan matematis dihitung intensitas atau frekuensi keberadaan kata. Metode LSA mengaplikasikan teori Singular Value Decomposition (SVD), yaitu teknik yang digunakan untuk melakukan estimasi terhadap rank dan matriks. Melalui SVD akan dilakukan peleburan atau reduksi dari matriks berukuran besar menjadi matriks kecil. Kemiripian antara kalimat dihitung dengan melakukan penghitungan nilai kosinus dari sudut atau membandingkan norma sudut Frobenius antara dua vector. Pada skripsi ini, ditambahkan program pengenalan kata frasa dan negasi dari sistem Simple-O sebelumnya yang kemudian dibandingkan dengan penilaian secara manual (human raters). Untuk menghitung nilai korelasi antar sistem dan human raters, maka digunakan teknik penghitung korelasi Pearson Product Moment. Nilai korelasi yang didapatkan antara sistem baru dan manual adalah 0.53155. ......Simple-O is an automated essay scoring system (essay grading) that applies Latent Semantic Analysis (LSA) method. The working principle of this system is by converting the answers into the form of a matrix which is then statistically and mathematically calculated the intensity or frequency of the existence of the word. LSA method applying Singular Value Decomposition theory (SVD), which is a technique used to estimate the rank and matrix. SVD will be done through reduction of a large matrix into smaller matrices. The similarity between the sentence is calculated by calculating the value of the cosine of the angle or comparing Frobenius angles norm between two vectors. In this paper, the program added phrases and negation words identification of Simple-O system and will be compared with Original Simple-O and human raters. To calculate the value of the correlation between the system and human raters, we used the technique Pearson Product Moment. Correlation values obtained between the new system and the manual at about 0.53155

Abstrak :
Sistem Penilaian Esai Otomatis (SIMPLE-O) dikembangkan oleh Departemen Teknik Elektro Fakultas Teknik Universitas Indonesia untuk ujian bahasa Indonesia. Skripsi ini akan membahas mengenai pengembangan SIMPLE-O untuk penilaian ujian bahasa Indonesia menggunakan metode Siamese Manhattan Long Short-Term Memory (LSTM) dan bahasa pemrograman Python. Terdapat dua dokumen yang akan menjadi input, yaitu jawaban esai dari peserta ujian dan jawaban referensi dari penguji. Kedua jawaban diproses dengan layer LSTM yang sama. Selanjutnya, kemiripan antara keduanya dihitung dengan fungsi persamaan. Pengujian dengan dataset jawaban dummy mendapatkan nilai MAE dan RMSE sebesar 0.0254 dan 0.0346. Kemudia, pengujian dengan dataset jawaban asli mendapatkan nilai MAE dan RMSE terbaik sebesar 0.1596 dan 0.2190. Rata-rata nilai akurasi yang didapatkan adalah 92.82 untuk fase training dan 84.03 untuk validasi.

---

The Automatic Essay Assessment System (SIMPLE-O) was developed by the Department of Electrical Engineering, Faculty of Engineering, University of Indonesia for the Indonesian language test. This thesis will discuss the development of SIMPLE-O for the assessment of Indonesian language tests using the Siamese Manhattan Long Short-Term Memory (LSTM) method and the Python programming language. There are two documents that will be input, essay answers from examinees and answer answers from examiners. Both answers are processed with the same LSTM layer. Next, the similarity between the two is calculated by the similarity function. Testing with dummy answer dataset produces MAE and RMSE values of 0.0254 and 0.0346. Then, testing with the real answer dataset produces MAE and RMSE values of 0.1596 and 0.2190. The average accuracy value obtained was 92.82 for the training phase and 84.03 for validation.

Abstrak :
Skripsi ini membahas mengenai pengembangan Sistem Penilaian Esai Otomatis (SIMPLE-O) yang dirancang dengan menggunakan hybrid CNN dan Bi-LSTM dan Manhattan Distance untuk penilaian esai Bahasa Jepang. Sistem dirancang dengan menggunakan bahasa pemrograman Python. Sistem melalui tahapan pre-processing, feature extraction dan word embedding yang dilanjutkan dengan proses deep learning serta pengukuran dengan menggunakan manhattan distance. Hasil akhir dari sistem dibandingkan dengan penilaian manual oleh dosen. Model yang paling stabil dan terbaik ditraining dengan menggunakan hyperparameter dengan kernel sizes bernilai 5, jumlah filter atau output CNN sebesar 64, pool size sebesar 4, Bidirectional LSTM units 50, batch size sebesar 64. Model deep learning ditraining dengan menggunakan optimizer Adam dengan learning rate 0,001 , epoch sebanyak 25 dan menggunakan regularizer L1 sebesar 0,01. Rata-rata error yang diperoleh adalah 29%......This thesis discusses the development of an Automatic Essay Grading System (SIMPLE-O) designed using hybrid CNN and Bidirectional LSTM and Manhattan Distance for Japanese essay grading. The system is designed using Python programming language. The system goes through the stages of pre-processing, feature extraction and word embedding followed by deep learning process and measurement using Manhattan Distance. The final result of the system is compared with manual assessment by lecturers. The most stable and best model is trained using hyperparameters with kernel sizes of 5, number of filters or CNN outputs of 64, pool size of 4, Bidirectional LSTM units of 50, batch size of 64. The deep learning model is trained using the Adam optimizer with a learning rate of 0.001, epoch of 25 and using an L1 regularizer of 0.01. The average error obtained is 29%.

Abstrak :
Grading is a process for decision making using information from evaluation of learning result whether using a test instrument or not[1]. Grading with essay is on option to evaluate level of knowledge of the students, but essay grading is not giving an objective view to each student. Essay grading by many of researcher is considered a good tools to evaluate result of a learning process and so, to evaluate level of intuition like synthesis and analysis. [2]. This research is intended to create an automatic essay grading which is called SIMPLE (SIsteM PeniLaian Esei otomatis) using Latent Semantic Analysis (LSA) as one of the method to extract and represent sentence using mathematical calculation or statistic from large amount of text [3]. Mathematical calculation is done by mapping with or without word from matrix group of word Furthermore, this research is done by implementing weight feature on web based automatic essay grading using Indonesian language. Testing is done by comparing result from system that using weight word and system that not using weight word Testing has succeeded with 82.56-96.42 percentage agreement with human raters for system using weight word.