Luthfi Faisal Rafiq
Fakultas Ilmu Komputer, Universitas Brawijaya

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Pencarian Teks Pada Terjemahan Hadits Shahih Bukhari Dengan Metode WIDF dan Bray-Curtis Distance Luthfi Faisal Rafiq; Muhammad Tanzil Furqon; Marji Marji
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 4 No 12 (2020): Desember 2020
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

With the development of the application technology for searching the Hadith translation text in digital form, it is one of the good learning media to help the Hadith learning process. There have been many types of digital Hadith that have been spread. However, most of the Hadith translation text searches still use keyword-based search technology, which will result in some related Hadith translation texts not being displayed or irrelevant Hadith translation texts being displayed. To solve this problem, this study uses a term weighting algorithm and distance measurement to maximize the search results for the Hadith text. Several stages in this research are the text preprocessing stage which produces terms to represent documents, the term Weighted Inverse Document Frequency (WIDF) weighting method to provide weight values ​​for each term and the Bray-Curtis distance measurement to determine the similarity between the query and the document. In the testing process used 5 different queries and Mean Average Precision (MAP). The results of testing the effect of using term weighting, the term weighting with the best performance is WIDF weighting with a MAP value of 0.71 when the number of documents returned is 5. For the test results of the distance measurement algorithm, the best distance measurement algorithm is Bray-Curtis Distance with a MAP value of 0.71 while in Euclidean the MAP value is 0.2 on the same number of returned documents. Meanwhile, the results of the test evaluation showed that the best variation in the value of n on the number of n documents returned was 5 with a MAP value of 0.632. The greater the n value, the greater the diversity of documents, as a result, the greater the chance for irrelevant documents to be taken which causes the MAP value to decrease.