Iriani, Putri Rahma (2019) Temu Kembali Informasi Lintas Bahasa Untuk Dokumen Berita Berbahasa Indonesia-Inggris Menggunakan Metode Bm25. Sarjana thesis, Universitas Brawijaya.
Abstract
Berita merupakan kebutuhan informasi seseorang untuk mengetahui apa yang sedang terjadi. Usaha mendapatkan berita relevan dari keberagaman bahasa dan dokumen itu tidak mudah didapatkan. Dokumen berita biasanya ditulis dalam bahasa asing. Hal ini menjadi sulit karena tidak semua pengguna mengerti bahasa asing, sedangkan berita yang diinginkan terdapat pada kumpulan bahasa asing tersebut. Pengguna bisa saja membacanya satu persatu untuk memperoleh berita sesuai kebutuhan, tetapi proses tersebut tidak efisien dan akan membutuhkan waktu yang lama. Sistem pencarian berita otomatis lintas bahasa sangat dibutuhkan untuk menyelesaikan masalah ini, yang mana pengguna hanya memasukkan query dengan bahasa Indonesia atau Inggris dan sistem akan mengembalikan dokumen dengan kedua bahasa tersebut. Permasalahan tersebut dapat dipecahkan dengan membangun sistem untuk memperoleh berita secara otomatis tanpa terhalang faktor bahasa. Sistem ini dibangun menggunakan metode BM25 yang mana telah dibuktikan mampu mengembalikan dokumen relevan dengan pemeringkatan. Parameter bebas yang digunakan k1 = 2,5 dan b=8,0. Pembobotan dilakukan dengan memperbandingkan IDF BM25 dan IDF BM25 Modifikasi yang mana menghasilkan akurasi tertinggi dengan nilai sama yaitu sebesar 0,95 dengan nilai k=5 pada pengujian precision@k.
English Abstract
News is an information about someone's needs to find out what is happening. Efforts to get relevant news from a variety of languages and documents are not easy to obtain. News documents usually written in foreign language. This becomes difficult because not all users understand foreign language, while the news needed in the collection of foreign language. Users can read one by one to get news as it needed, but this process is inefficient and will take a long time. A cross-language automatic news search system is needed to solve this problem, where users only enter requests with the native language and the system will recover documents in other languages. This problem can solve by creating a system to obtain automatic news without language barriers. This system will builds using the BM25 method which has been proven to be able to improve documents that are relevant to the ranking. The free parameters used are k1 = 2.5 and b = 8.0. Weighting is done by comparing IDF BM25 and IDF modification which results in the highest value of 0.95 with k = 5 in testing of precision@k.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/FILKOM/2019/173/051902343 |
| Uncontrolled Keywords: | News is an information about someone's needs to find out what is happening. Efforts to get relevant news from a variety of languages and documents are not easy to obtain. News documents usually written in foreign language. This becomes difficult because not all users understand foreign language, while the news needed in the collection of foreign language. Users can read one by one to get news as it needed, but this process is inefficient and will take a long time. A cross-language automatic news search system is needed to solve this problem, where users only enter requests with the native language and the system will recover documents in other languages. This problem can solve by creating a system to obtain automatic news without language barriers. This system will builds using the BM25 method which has been proven to be able to improve documents that are relevant to the ranking. The free parameters used are k1 = 2.5 and b = 8.0. Weighting is done by comparing IDF BM25 and IDF modification which results in the highest value of 0.95 with k = 5 in testing of precision@k, BM25, cross language information retrieval, news |
| Subjects: | 000 Computer science, information and general works > 025 Operations of libraries, archives, information centers > 025.5 Services for users > 025.52 Reference and information services > 025.524 Information search and retrieval |
| Divisions: | Fakultas Ilmu Komputer > Teknik Informatika |
| Depositing User: | Nur Cholis |
| Date Deposited: | 15 Jun 2020 13:45 |
| Last Modified: | 24 Oct 2021 04:42 |
| URI: | http://repository.ub.ac.id/id/eprint/169095 |
Preview |
Text
Putri Rahma Iriani (2).pdf Download (2MB) | Preview |
Actions (login required)
 |
View Item |