Voice Assistant for WEGO

No Thumbnail Available

Date

2023

Journal Title

Journal ISSN

Volume Title

Publisher

Library Information Services, CUI Lahore

Abstract

This project has been developed as a part of an industrial FYP with WEGO. WEGO is a travel assistant application that allows the user to search for and book flights and hotels. We were assigned the task of automating the process of searching the flight through WEGO with the help of a voice assistant. This project will improve the user experience of searching for a flight through WEGO and make the user less prone to errors. The voice assistant for the project has been trained using Google’s Dialogflow and the application has been developed using Kotlin and Jetpack Compose. This is a standalone Android application and has not been integrated into the existing WEGO application. Moreover, the application provides a complete chatbot experience where the user can communicate with the voice assistant regarding the flight they want to search. Once the user has communicated all of the required information with the voice assistant, they will be shown their search results on a WebView page from WEGO within the application. Since the majority of WEGO’s user base lies in the middle east, we have added Arabic and Urdu language support to the application as well. Further functionalities of our application include user accounts and user history. If the user is registered with the application, their history will be saved and they can visit their old search results from the history. User accounts are maintained using Firebase authentication and history is maintained using Firebase Realtime Database. We have succeeded at solving the problem that we set out to solve with this application. The application is working as it should and showing the relevant results. As far as the accuracy of the search results is concerned it is 100% when it comes to taking input in the English language. However, the accuracy is slightly compromised when it comes to Urdu and Arabic. We have used Google’s on-device translation models for adding Urdu and Arabic languages which compromises the accuracy of the search results shown. However, the search results are still respectable.

Description

Keywords

Voice Assistant for WEGO

Citation

Endorsement

Review

Supplemented By

Referenced By