Sistem Pendeteksi Dokumen Plagiat Harfiah pada Dokumen Teks Berbahasa Indonesia dengan Memanfaatkan Mesin Pencari

Abstract

Search engines can be used to detect plagiarism because search engines are one of the gateways to get source documents. This research aims to establish a corpus of document plagiarism and develops a system that can detect plagiarism by utilizing search engines. The corpus is created by copying passages from 1-3 source documents and restructuring the source documents by translating back and forth with Google Translate. The corpus consists of 100 documents. The documents are extracted into segments consisting of 4-20 words. The segments will be weighted based on the words existence in Indonesian dictionary where words not found in dictionary are given higher weights. Using Google’s search engine, this study successfully detects 100% of the plagiarized documents using only a maximum of 31% segments. On the other hand, using Bing and 40% segment documents only detects 30% of the corpus. The results of this study show that the performance of online plagiarism detection depends on the quality of the search results provided by search engines