DETEKSI KEMIRIPAN TOPIK PROPOSAL JUDUL TUGAS AKHIR DAN SKRIPSI MENGGUNAKAN LATENT SEMANTIC ANALYSIS DI STMIK BUMIGORA MATARAM

  • I Putu Hariyadi
  • Hartarto Junaedi
Keywords: latent semantic analysis, enhanced confix stripping stemmer, term weighting, similarity detection, cosine similarity, mean average precision

Abstract

Research in university has important role in contributing to national development. By knowing the importance of research, students are motivated to be involved in a research which makes contribution to science. Therefore, the appropriateness of research topic taken by students need to be verified. The result of manual verification process is neither efficient, effective, nor accurate. Thus, methods employing Information Technology (IT) are being developed nowadays. This research applied the Latent Semantic Analysis (LSA) method to detect similarity of topic research title. There are 4 steps in applying LSA method; those are preparation, preprocessing, similarity detection and evaluation step. The experimental result using 40 title proposal query for 400 undergraduate final assignments showed that this system is able to detect topic similarity of thesis title proposal with value MAP of 0.8465 on reduced value k=210 with Threshold Cosine similarity of > 0 without DF thresholding. Whereas testing by DF thresholding resulted in MAP value of 0.8744 on reduced value k=270 with threshold cosine similarity > 0.

References

[1] Christopher D. Manning, Prabhakar Raghavan, dan Hinrich Schutze, An Introduction to Information Retrieval, Cambridge University Press, 2009
[2] Cios Krzysztof J., Witold Pedrycz, Roman W., Swiniarski, dan Lukasz A. Kurgan, Data Mining A Knowledge Discovery Approach. Springer, 2007.
[3] Dr. E. Garcia, Document Indexing Tutorial for Information Retrieval Students and Search Engine Marketers, Mi Islita.com, 2005.
[4] Oystein Lohre Garnes, Feature Selection for Text Categorisation, Norwegian University Science and Technology, 2009
[5] Thomas K Landauer, Peter W. Foltz, dan Darrell Laham, “An Introduction to Latent Semantic Analysis”, Discourse Processes, hal. 259-284, 1998
[6] Ria Hari Gusmita, dan Ruli Manurung, "Penerapan Latent Semantic Analysis (LSA) untuk Menentukan Kesamaan Makna antara Kata dalam bahasa Inggris dan Kata dalam Bahasa Indonesia", Universitas Mercu Buana, hal.8, 2008
[7] Baker, Kirk. Singular Value Decomposition Tutorial. The Ohio State University, hal. 15, 2013
Published
2016-10-29
Section
Articles