PENGEMBANGAN APLIKASI "LOST & FOUND" BERBASIS ANDROID DENGAN MENGGUNAKAN METODE TERM FREQUENCY – INVERSE DOCUMENT FREQUENCY (TF-IDF) DAN COSINE SIMILARITY

Abstract

Abstract
The search process will generally be carried out by spreading information from person to person, distributing information papers in public places, reporting to authorities and also through information media such as newspapers, radio, television and social media. The process of searching through information media is certainly very effective, because the more people who know it, the more likely it is to be found. But this method still has weaknesses because the goods that have not been found to even take months and even years, the loss of information will be covered with new news developing in the community so that the loss of information that has been made will begin to be forgotten by many people. Against this background, the researchers encouraged researchers to develop the "Lost & Found" Application. The application can be accessed via android mobile online. The purpose of developing the "Lost & Found" application is to prepare a forum for individuals to disseminate lost information or findings. The application is implemented using the Cosine Similarity Method, which is useful for finding how similar the level of similarity between documents, the greater the cosine value, the more similar the documents are compared. A cosine value of 1 represents 100% similarity, while zero represents 100% dissimilarity [1]. The application will also apply the TF-IDF method where this method serves to give weight to the relationship of a word to the document.