Penerapan Algoritma Jaro-Winkler Untuk Autocorrect Dan Spelling Suggestion Pada Aplikasi Speech Recognition CMS Berbasis Website

Abstract

PT. Bukaloka Teknologi Indonesia is an online buying and selling site that distributes products and services equipped with an online store & instant web generation platform where this service is a platform that integrates the marketplace with Online Company Profile creation services for micro, small, and medium enterprises (UMKM) who joined. Currently, the process of changing UMKM Company Profile information still has to use a standard Content Management System (CMS) which requires changes to be made in the CMS dashboard itself and data entry is done by typing in the data form. This study discusses the implementation of Speech Recognition which is supported by the Jaro-Winkler Algorithm which can simplify and speed up the process of entering words compared to using CMS without the need for the use of a keyboard. The Jaro-Winkler algorithm itself is intended to improve words and the process of word accuracy on potential word errors generated by Speech Recognition. 6 prospective users were surveyed to determine the feasibility of the application in the community and get the results of the survey calculations with the number of responses saying "Yes" as many as 87.20%, while "No" as many as 12.8%. With a total of 87.20%, it can be concluded that the result is "Successful".