Keywords Search Correction Using Damerau Levenshtein Distance Algorithm

Enny Dwi Oktaviyani, Sherly Christina, Deddy Ronaldo

Abstract

Searching is one of the important features on the website, but it is not uncommon for users to make typos when typing keywords. Typing errors of these keywords is usually referred to as typo. This study aims to build a system by providing suggestions for correcting typos in the search feature. Keywords search correction are obtained using the Damerau-Levenshtein Distance Approximate String Matching algorithm by to calculate the editing distance of each word in a keywords with each word in the Indonesian word dictionary. Testing was carried out as many as 40 experiments, with 10 keywords and 250 articles taken randomly. The test results show the Damerau-Levenshtein Distance algorithm is able to provide precision and recall values of 91.24% and 89.58% in providing keyword improvement suggestions. With the improvement of the system, each trial increases with precision value of 0.80 and recall value of 0.98

Keywords

Damerau-Levenshtein Distance, searching, keywords, correction

References

Maghfira, T. N., Cholissodin, I., & Widodo, A. W. (2017). Deteksi Kesalahan Ejaan dan Penentuan Rekomendasi Koreksi Kata yang Tepat Pada Dokumen Jurnal JTIIK Menggunakan Dictionary Lookup dan Damerau-Levenshtein Distance. Malang. Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer, 1(6), 498-506.

Mishra, R., & Kaur, N. (2013). A survey of spelling error detection and correction techniques. International Journal of Computer Trends and Technology, 4(3), 372-374.

Sutisna, U., & Adisantoso, J. (2010). Koreksi Ejaan Query Bahasa Indonesia Menggunakan Algoritme Damerau Levenshtein. Jurnal Ilmiah Ilmu Komputer, 8(2).

Jogiyanto, H. M. (2008). Metodologi penelitian sistem informasi. Yogyakarta: Andi Offset.

Nugroho, A. (2009). rekayasa perangkat lunak menggunakan UML dan JAVA. Penerbit Andi.

Wibowo, A. (2011). Pengujian Kerelevanan Sistem Temu Kembali Informasi. In Seminar Nasional Ilmu Komputer.

Article Metrics

Abstract view : 9 times

Refbacks

  • There are currently no refbacks.