Pengelompokan Tematik Al-qur'an Menggunakan Metode LSA dengan Pembobotan Log Normalization dan IDF

Sarowis, Maulana (2020) Pengelompokan Tematik Al-qur'an Menggunakan Metode LSA dengan Pembobotan Log Normalization dan IDF. Other thesis, Universitas Islam Riau.

Text 153510717.pdf - Submitted Version Download (9MB)

Abstract

The Qur'an is a revelation revealed by Allah SWT to the Prophet Muhammad SAW through the intermediary Angel Gabriel which in the Qur'an consists of 30 Juz (parts), 114 surahs and about 6236 verses. Verses in the Qur'an have links between one verse with another and various kinds of discussion. Problems that occur in order to get a discussion that is desirable for example about the theme of heaven the need for commentators, hadith experts and other supporters. This research aims to improve the efficiency of scientific search and reduce the difficulty in grouping surahs and verses. In this study the dataset used is, the translation of the Qur'an datasetqurandatabase.org. This grouping is done using the method Latent Semantic Analysis (LSA) by weighting Log Normalization and IDF which is an algebraic statistical method that extracts the hidden semantic structure from words and sentences. This study produced a grouping with subthemes "The Nature of Heaven and Enjoyment" and "Immortality of Heaven" which obtained the value of f1-score evaluation of the Cordoba Al-Qur'an on average 62.64% of the two matrix evaluations VT and U, whereas from experts the thematic Al-Qur'an the contribution of adding 2 surahs and verses to the sub theme "The Nature of Heaven and Enjoyment" with an increase in f1-score of 1.64%. Keywords: Al-Qur’an, Cordoba, Latent Semantic Analysis(LSA), Log Normalization, Invers Document Frequency(IDF)

Actions (login required)

View Item