Abstract :
News serves as a medium of information used to acquire various results of information. The information conveyed through news can influence readers' understanding and knowledge regarding specific events or topics. The issue that has arisen this year is the political year, where there has been a significant prevalence of unethical campaign practices known as black campaigns. A black campaign is an unlawful act in political campaigning ethics, involving the spread of false news and the undermining of political opponents with negative narratives. If such practices persist, they can lead to unrest and conflict between groups. Therefore, it is crucial to identify false news to combat black campaigns. There are independent website channels for identifying false news, such as CEKFAKTA. However, in identifying false news, users must manually re-verify, which is time-inefficient. Consequently, there is a need for efficiency in identifying false news without the necessity of re-verification. Therefore, this research employs a text similarity approach using the cosine similarity algorithm to expedite the identification of false news from black campaigns. The dataset used consists of 300 fake and real news items. Based on the test results, an accuracy rate of 87% was achieved, leading to the conclusion that the cosine similarity algorithm is capable of identifying false news.