KLASIFIKASI TEKS PORNOGRAFI BERBASIS MACHINE
LEARNING MENGGUNAKAN METODE NAIVE BAYES
Total View This Week1
Total View This Week1
Institusion
STMIK Bumi Gora
Author
HENDRI, JAYADI
Subject
TN Mining engineering. Metallurgy
Datestamp
2021-09-01 06:57:07
Abstract :
The spread of information through the internet is quite fast. Not only has the positive impact, the internet also has a negative impact, one of which is the ease of finding information that contains negative things such as pornographic stories. Based on this review, the purpose of this research is how to prevent early content containing the word pornography based on the title of the content on the web page. The prevention process is carried out by classifying using the naive Bayes method and the weighting process with TF IDF and without TF IDF. The type of data used is in the form of content titles that have been downloaded via internet sites such as social media and web pages. There are 200 story title datasets that are used. The results of the test using the Data Splitting method with TF-IDF values and without TF IDF get Accuration, Precision, and Recall up to 76.44%, 82.43%, 73.43%. And without using TF-IDF, you get Accuration, Precision, Recall up to 97%, 96%, and 93%, respectively. The conclusion is that the Naïve Bayes method without TF-IDF creates higher accuracy than using TF-IDF.
The spread of information through the internet is quite fast. Not only has the positive impact, the internet also has a negative impact, one of which is the ease of finding information that contains negative things such as pornographic stories. Based on this review, the purpose of this research is how to prevent early content containing the word pornography based on the title of the content on the web page. The prevention process is carried out by classifying using the naive Bayes method and the weighting process with TF IDF and without TF IDF. The type of data used is in the form of content titles that have been downloaded via internet sites such as social media and web pages. There are 200 story title datasets that are used. The results of the test using the Data Splitting method with TF-IDF values and without TF IDF get Accuration, Precision, and Recall up to 76.44%, 82.43%, 73.43%. And without using TF-IDF, you get Accuration, Precision, Recall up to 97%, 96%, and 93%, respectively. The conclusion is that the Naïve Bayes method without TF-IDF creates higher accuracy than using TF-IDF.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
