Semantic-Based Search Design For Staffing Data Using Cosine Similarity Method

Abstract

PT. Kenari Ketchup is one of the suppliers and distributors of the Kenari Ketchup brand. Employees at PT. Kecap Kenari has different tasks, educational backgrounds, and expertise. Researchers made observations at PT. Kecap Kenari, it was found that the search for employee data was still manual based on keywords only. To overcome this, a feature is needed that allows data search besides being based on keywords, also based on the semantic similarities/meanings of keywords. So that it can support data search with a more complete search scope. Some of the main stages in searching for data in this study are the text preprocessing stage, the search for keyword semantic meanings, and the string matching stage. At the preprocessing stage the text utilizes the PHP Literary library to normalize the text, the semantic meaning search stage uses the PHP Satria/Synonim Antonym library, while the employee data search uses the Cosine Similarity method as a string matching technique to determine the level of similarity between employee data and keywords and the semantic meaning of the keywords used . The results of this study, that the system succeeded in normalizing the text, finding the semantic meaning of the keywords, and finding the level of similarity of the data sought with cosine similarity, as well as presenting data sorted from the highest level of similarity, so as to make it easier for users to read system output related to searching employee data