Ananda P, Diajeng Tania (2019) Prediksi Rating Otomatis Berdasarkan Review Restoran Pada Aplikasi Zomato Dengan Menggunakan Extreme Learning Machine (ELM). Sarjana thesis, Universitas Brawijaya.
Abstract
Seiring berkembangnya zaman, perkembangan teknologi semakin berkembang. Salah satunya aplikasi untuk mencari informasi mengenai restoran di Jakarta yaitu Zomato. Zomato merupakan aplikasi yang menyediakan informasi dari berbagai restoran, fasilitas, harga, review, dan rating dari restoran tersebut. Masyarakat dapat memberikan review dan rating pada restoran tersebut. Data review tersebut berguna untuk pengguna sebelum ke restoran tersebut. Data review tersebut terkadang belum disertai rating sehingga membuat pemilik restoran mendapatkan masalah dalam mengklasifikasikan review kedalam rating untuk melakukan evaluasi kedepannya pada restoran tersebut. Pada penelitian ini membantu untuk mengklasifikasikan review kedalam rating. Pengujian metode ini menggunakan prediksi dengan metode Extreme Learning Machine (ELM). Proses prediksi ini menggunakan tahapan pre-processing, pembobotan kata dengan TF-IDF, dan perhitungan metode Extreme Learning Machine (ELM). Terdapat tahapan-tahapan pada metode ELM antara lain normalisasi, proses training, dan proses testing. Metode ELM ini menghasilkan akurasi sebesar 80,01% Dengan jumlah k yaitu 10 menggunakan hidden neuron sebanyak 25 dengan Interval bobot -0,5 hingga 0,5 fungsi aktivasi Sigmoid biner. Dapat disimpulkan bahwa metode ELM dapat menyelesaikan masalah prediksi dengan cukup baik.
English Abstract
In this modern culture, technology advancement are growing better than we ever discovered before. One of the apps we use to search for information about restaurant in Jakarta are known as Zomato. Zomato is an application that provides various information about a restaurant from it facility, price, review, and rating. Users of The Zomato App can input various information that people haven’t aware of about the restaurant into the app. Besides of inputting information into the app, Users of The Zomato App can also input a review and rating of a specific restaurant. The data review is used as an information about the restaurant for the potential customer from The Zomato App but sometimes the data review doesn’t yet include a restaurant rating. This lack of misinformation will surely make the restaurant owner to occure some difficulties such as improving the restaurant services status for future outcomes. This research helps to classifying the review into the rating. Test protocol of this research are using a prediction with Extreme Learning Machine (ELM) Methods as it core. The prediction process however are build from a several steps such as pre–processing, word weighting with TF-IDF, and Extreme Learning Machine (ELM) Method calculations. Test result of The ELM parameter provides accuracy result 80,01% with k=10 amount hidden neuron 25 Interval weights -0.5 until 0,5 using function activation Sigmoid biner. We have come to conclusion were ELM method could positively solve the prediction problem exquisitely.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/FILKOM/2019/111/051902281 |
| Uncontrolled Keywords: | prediksi rating, review, Extreme Learning Machine-Rating prediction, Review, Extreme Learning Machine |
| Subjects: | 000 Computer science, information and general works > 006 Special computer methods > 006.3 Artificial intelligence > 006.31 Machine learning |
| Divisions: | Fakultas Ilmu Komputer > Teknik Informatika |
| Depositing User: | soegeng sugeng |
| Date Deposited: | 27 Jul 2020 07:55 |
| Last Modified: | 19 Oct 2021 09:02 |
| URI: | http://repository.ub.ac.id/id/eprint/168895 |
Preview |
Text
Diajeng Tania Ananda P.pdf Download (2MB) | Preview |
Actions (login required)
 |
View Item |