OPTIMIZATION OF NAÏVE BAYES USING LEVENSHTEIN DISTANCE FOR TYPOGRAPHICAL ERROR CORRECTION IN SENTIMENT ANALYSIS

Utomo, Aziz Yogo (2022) OPTIMIZATION OF NAÏVE BAYES USING LEVENSHTEIN DISTANCE FOR TYPOGRAPHICAL ERROR CORRECTION IN SENTIMENT ANALYSIS. S1 - Sarjana thesis, Universitas AMIKOM Yogyakarta.

	Text (COVER-ABSTRAK) COVER.pdf Download (2MB)
	Text (BAB I) BAB I.pdf Download (291kB)
	Text (BAB II) BAB II.pdf Restricted to Registered users only Download (632kB)
	Text (BAB III) BAB III.pdf Restricted to Registered users only Download (291kB)
	Text (BAB IV) BAB IV.pdf Restricted to Registered users only Download (554kB)
	Text (BAB V) BAB V.pdf Download (44kB)
	Text (DAFTAR PUSTAKA) Daftar Pustaka.pdf Restricted to Registered users only Download (147kB)
	Archive (SOURCE CODE) Source Code-18.61.0149-Aziz Yogo Utomo.zip Restricted to Repository staff only Download (62kB)
	Text (PUBLIKASI) Publikasi-18.61.0149-AzizYogoUtomo.pdf Restricted to Repository staff only Download (413kB)

Abstract

Now we live in an era with a tremendous amount of unstructured data such as text data. Naïve Bayes is an algorithm that is well-performed for dealing with text data. In processing text data, there are several problems, such as the vast amount and the data dimension. Text data derived from human fingers allows typographical errors in writing; this typographical error becomes another problem because it will make the data dimension bigger and change the semantics of the word itself. In addition, typographical errors liable the calculation of the Naïve Bayes Likelihood to be 0 and affect the performance of the model. Levenshtein Distance is one method to correct typographical errors. This research indicates that the Levenshtein Distance has succeeded in increasing the performance of the Naïve Bayes model with optimal results using distance 2 of 5,9%

Actions (login required)

View Item