The title classification system aims to help users to Group A student seminar title document into a Category. The number of Student Research title documents requires an application system that can group these documents according to their respective categories. The purpose of this study is how to build a seminar title classification system by utilizing the approach through the text of the title and description of the seminar with the TF-IDF Method to give weight to the frequency of the relationship on the occurrence of a word (term) in the calculation of the merger of title and description of the seminar. And the Cosine Similarity algorithm is used as a comparison method to know how much similarity between the two documents. This research is conducted by implementing the text mining method with cosine similarity algorithm and TF-IDF weighting so that it is expected to classify data automatically, quickly, and accurately. Using dummy data as training data used in this study, totaling 100 seminars that have been carried out with several categories of different fields of science. And for its implementation using the PHP programming language, and with the help of data resources the use of text mining that is already in PHP which is useful for the text preprocessing stage, and then will continue with the implementation process of TF-IDF Motode and Cosine Similarity algorithm.
Copyrights © 2024