Implementation of Algorithm Rabin-Karp for Thematic Determination of Thesis

Abstract

Thesis / scientific work is a science composition that presents facts and is written according to a good and correct writing methodology. But to find out what the theme discussed should read the abstract of the essay. Along with the development of information technology that offers convenience for human labor in simplification of his work. Difficulty in determining the theme in the thesis can be aided with applications that are able to set themes based on abstract. To make the application of thesis thematic determination can use string matching algorithm, one of string matching algorithm is Rabin-Karp algorithm by doing string matching based on the hash value on text and hash value in a pattern. The Rabin-Karp algorithm is often used to detect plagiarism text document, comparing the overall hash value of k-gram contained in all text documents, then deciding the degree of similarity. Then it can be concluded that Rabin-Karp is able to string a text as a whole, from this the researcher is interested in proving what if the Rabin-Karp algorithm is implemented in the classification field wherein in this study it is used to string matching each word or keyword refers to the theme of an Indonesian essay or thesis abstract. The application is built using web programming language PHP and MySQL database. The confusion matrix is used to determine accuracy, obtained 88% and 12% for misclassification rates.