Pengembangan Sistem Jurnal dan Penggunaan Text Mining Cosine Similarity dan SVM untuk Penilaian Plagiarism pada Judul dan Abstrak
Total View This Week0

Institusion

Institut Teknologi Perusahaan Listrik Negara

Author

Mandana, Rizal Adi
Sudirman, M. Yoga Distra
Kusuma, Dine Tiara

Subject

Teknik Informatika 

Datestamp

2022-09-23 02:00:07 

Abstract :
Early analysis of journal plagiarism is needed by the editorial team and partners to evaluate the authenticity of the proposed journal, assessment of plagiarism is needed to maintain the quality of journal publications. Therefore, a plagiarism analysis feature is needed in journal management applications, so it will help in providing information about the journal's authenticity. In this study SDLC Waterfall is used by applying the Text Mining Cosine Similarity method and SVM in assessing the level of similarity of a proposed journal with existing journal data. Application design uses UML and blackbox testing is conducted to determine the success of applications and evaluation methods to determine the accuracy of document similarity classification. The results of plagiarism testing on the electronic journal portal STT-PLN on 30 published journal journal data with 1836 training data using the Text Mining Cosine Similarity method and SVM can identify early plagiarism in journals with 100% precision and 100% recall with susceptible results plagiarism values 5,52% - 29,03%, but if the amount of training data big then takes a long time. With the early analysis of plagiarism can improve the quality of journal publications 

File :
Skripsi_Rizal Adi Mandana(201431067).pdf