Ahmad Zainal Abiddin, - (2018) PENGEMBANGAN SISTEM DATA-TO-TEXT (D2T) UNTUK MEMBANGKITKAN BERITA PADA DATA STREAMING. S1 thesis, Universitas Pendidikan Indonesia.
![[img]](http://repository.upi.edu/style/images/fileicons/text.png) |
Text
S_KOM_1404862_Title.pdf Download (18kB) |
|
Text
S_KOM_1404862_Abstract.pdf Download (126kB) |
|
Text
S_KOM_1404862_Table_of_content.pdf Download (71kB) |
|
Text
S_KOM_1404862_Chapter1.pdf Download (109kB) |
|
Text
S_KOM_1404862_Chapter2.pdf Restricted to Staf Perpustakaan Download (835kB) |
|
Text
S_KOM_1404862_Chapter3.pdf Download (193kB) |
|
Text
S_KOM_1404862_Chapter4.pdf Restricted to Staf Perpustakaan Download (2MB) |
|
Text
S_KOM_1404862_Chapter5.pdf Download (64kB) |
|
Text
S_KOM_1404862_Bibliography.pdf Download (215kB) |
|
Text
S_KOM_1404862_Appendix.pdf Restricted to Staf Perpustakaan Download (2MB) |
Abstract
Penelitian ini bertujuan untuk mengembangkan sistem Data-to-Text dengan masukan berupa data Streaming dalam bentuk batch, untuk membangkitkan berita secara general. Pengembangan model sistem Data-to-Text dilakukan penerapan Machine Learning untuk mengatasi permasalahan data Streaming, dengan teknik Piecewise Linear Approximation menggunakan Least Square method. Sistem yang dikembangkan menghasilkan informasi ringkasan data, informasi data terkini, dan informasi prediksi. Pengembangan sistem dilakukan dalam bahasa pemrograman R dengan memanfaatkan beberapa packages yang tersedia. Eksperimen dilakukan dengan mengukur tingkat Readability dari berita yang dibangkitkan, Computation Time, dan membandingkan hasil dengan penelitian terkait. Hasil eksperimen menunjukan bahwa informasi yang dihasilkan terbukti merepresentasikan data yang diberikan, dan dapat dipahami oleh tingkat mahasiswa atau diatasnya, serta waktu komputasi cukup baik. Sistem ini mampu menghasilkan informasi berdasarkan data meteorologi, data klimatologi, dan data keuangan.;---The study aims to develop a Data-to-Text system with input streams data in the form of batches, to generate general news. Development of a Data-to-Text system model is done by applying Machine Learning to resolve data streaming problem, with Piecewise Linear Approximation technique using Least Square Method. This system produces a summary text of data, current data description, and predict information. System development is done in R programming language by utilizing several available packages. Experiments were carried out by measuring the level of Readability of news that was raised, Computation Time, and compared with related research. The experiment results show that generated news proven to represent data provided, can be understood by students with college level or above, and Computation Time is good. This system can generate a variety of news based on given data, like climates data, meteorology data, and exchange rate of money.
| Item Type: | Thesis (S1) |
|---|---|
| Additional Information: | No. Panggil : S KOM AHM p-2018; Pembimbing : I. Lala Septem, II. Enjang Ali; NIM. : 1404862. |
| Uncontrolled Keywords: | Data-to-Text, Natural Language Generation, Machine Learning, streaming, Picewise Linear Approximation, Least Square Method; Time-series, Data-to-Text, Natural Language Generation, Machine Learning, streaming, Picewise Linear Approximation, Least Square Method, Time-series. |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science T Technology > T Technology (General) |
| Divisions: | Fakultas Pendidikan Matematika dan Ilmu Pengetahuan Alam > Program Studi Pendidikan Ilmu Komputer |
| Depositing User: | Isma Anggini Saktiani |
| Date Deposited: | 02 Dec 2019 04:52 |
| Last Modified: | 02 Dec 2019 04:52 |
| URI: | http://repository.upi.edu/id/eprint/38305 |
Actions (login required)
 |
View Item |