Identifikasi Konten Hoaks Media Sosial tentang Covid-19 Menggunakan Algoritma Levenshtein Distance

Abstract

Social media is one of the developments in information technology. Where with the development of this technology, namely the internet, it can make it easier for all people to use it. There is a lot of information circulating, but not all the news that is informed is true. There have been various cases of spreading hoaxes or lies related to covid-19 which made people panic and immediately believe in the news that was spread even though the truth was unknown, whether it was a fact or a hoax. Therefore, a hoax news identification system was built using the Tf-Idf (Term frequency - Inverse document frequency) algorithm to measure the weight of a word in a document and the Levensthein Distance (LD) method. The application of the Levensthein Distance method in the hoax identification system has several stages including inputting data. Then proceed with the preprocessing stage including case folding, tokenezing, stopwords, and stemming. The next stage is the process of calculating word weights using the Tf-Idf (Term Frequency-Inverse Document Frequency) algorithm and the last stage is calculating the drinking distance between words using the Levensthein Distance method. The results of this study obtained an accuracy value of 0.87%, precesion 1.00%, recall 0.59% and f1-score 0.75% with a total dataset of 600.