Analisis Sentimen Situs Pembajak Artikel Penelitian Menggunakan Metode Lexicon-Based

Abstract

The development of technology is very rapid in today's era, especially with the development of internet technology. In the field of education itself, publication and access to research references can be easily done via the internet. However, with the internet also, research journal piracy can occur when access to references cannot be obtained due to paywall barriers (cannot access if you do not pay). Expressions of this can be found on social media, where users can use the platform to express their opinions freely. The availability of these data on the internet that can be retrieved and analyzed so as to produce information. In this study, we will analyze public sentiment towards one of the research article hijacking sites, namely Sci-Hub. The method used is Lexicon-based which is a method that uses a dictionary for classifying sentiment text. The data used are 35,755 tweets with 28,193 English-language tweets and 7,552 Indonesian-language tweets. The data is classified according to the rules in the InSet dictionary with the results of the classification of positive sentiment 59.51%, negative sentiment 26.42% and neutral sentiment 14.07%. This study obtained an accuracy of 55.66% for the prediction of data with positive, negative and neutral sentiments.