Utilizing Website-Based Speech-to-Text for Query Search

Abstract

Companies are increasingly using database-based data processing systems to store all information about their goods and products, so searching for queries on the system is one of the routine activities carried out by employees daily. The user must enter the keywords from the desired query in the search feature available on the system using the keyboard to perform the search process. It can only be accessed via a computer connected to the internet. Office However, sometimes there are several factors that make it difficult for employees to type in keywords, such as driving or having physical limitations, especially in the hands. This study aims to present an alternative to entering keywords not only through the keyboard but also by using voice by creating a query search system with the Speech-to-Text method so that users simply say the keywords from the desired data, and the system will be built based on a website so that they can be accessed from other devices such as smartphones, tablets, laptops or other devices. The method used in making this system is the System Development Life Cycle (SDCL) Prototype model and the system design modeling using the Unified Modeling Language (UML). This system will also be hosted on the Heroku platform, and in order to guarantee optimal operation of the system, a black box testing method is used, and the word accuracy test resulting from the Speech-to-Text conversion as well as the processing time required from the start of the system running to displaying the search results. It is hoped that with this system, employees will find it easier to enter keywords from the data they want through various devices quickly and with precise results.

GS Cited Analysis