Interactive Program to Prepare for The Actual Visa Interview.

Abstract

Now a day acquiring a VISA is getting tougher day by day and due to this reason most of the VISA applicants are getting rejected in the VISA interview because of lack of communication skills and no proper training. An interactive computer program would encourage user to get trainings for the interview. Also an interview training and guidelines may become expensive for the applicant.From an interactive program the user can get some idea and prepare accordingly for the actual VISA interview. This paper proposes a VISA Chatbot which will help the applicant train for the interview. Instead of traditional Chatbots which are rule-based bots, this Chatbot will be using machine algorithms for helping the VISA applicants. This Chatbot will reside in a Facebook page as it will be easier for the applicants to simply access the Facebook page, instead of downloading the Chatbot application.

Introduction

Chatbots are the next generation of programs that can interact with humans. Almost every chatbot has two components. First component is the message typed by the user or an audio message and the second component is the message output given by the chatbot. The whole process is repeated until user decides to end the interaction. The conversation happens between a human and a computer program in a programming way. For example, ELIZA was the first chatbot which take input keywords from user message and response it back accordingly. Chatbots can also help companies to gain revenue by providing advanced healthcare services.

Literature Review

Chatbots have been around since the mid-1960s. In 2007, Bayan talked about the use of chat bot for educational purpose. He said that ‘Chatbots could be used as a tool to learn or study a new language; a tool to access an information system, a tool to visualize the contents; and a tool to give answers to answers to questions in a specific domain…a chatbot could be trained with any text in any language’ [3]. So in 2013, Patrick and Bii followed Bayan’s statement and developed a chatbot technology whose main aim was to unlock the potential of students and use the chatbot for educational purpose. They used XML for storing questions and the response to the question.

Parab and Archana (2017) did a recent project on career counselling chatbot where the user will ask for career counselling from the bot which used the same NLP. Another similar and recent project was done by Cameron and Gillian (2017) where the chatbot was used in mental healthcare counselling purpose and developed into a full web based program.Polatidis, Nikolaos (2014) developed their bot who will take care of college admissions. Applicants will ask the admission bot about the college admission process and the bot will reply accordingly along with other information like accommodations, college fees and eligibility criteria.

Amilon and Mattias (2015) made a chatbot which used a common-sense database. A common-sense database is a where the information is stored that most people would take it as a common knowledge. For example, conceptnet was created as common-sense database. Their main aim was to check if the chatbot behavior can be improved using this common-sense database.

Purpose

Chatbot is one of the emerging technologies in field of data science. This project’s purpose is to make a Chatbot who will handle the VISA operations. So what is a Chatbot? A Chatbot is a program that communicates with a person. It is powered by Machine Learning Algorithms and applying AI concepts. The most famous example of a Chatbot is Apple’s Siri, Microsoft Cortana and Google Assistant. Their sole purpose is to help and reduce the manpower. Our Chatbot is going to do same thing, who will take VISA interview of the applicants.II. Product Scope We aim to build a Chatbot that will help many users to get VISA interview training with it and will reduce people’s time. It can be integrated with Facebook and it can also help people to get trained/prepared for their VISA interview. In short, we aim to ease people’s life.

Architecture of VISA Bot

The architecture consists of VISA bot, Heroku app, API call and NLP. The user using Facebook messenger will pass message to the bot. The bot will send the message to Heroku server where the bot actually resides. The API will be called along with the message and call the NLP. In NLP, it will take the message from user and perform identifying intents. Intents is where the keywords are taken and matched with the sentence. It will also start extracting entities where the work synonyms are stored and generate a response and send it back to user.

Use Case Diagram

The use-case diagram is shown in figure 2, where any user be it be an admin or tester or a normal user should be already logged in on Facebook. The user will then have to navigate to the VISA bot Facebook page and open the chat window to start the chat with the bot. He/she can also use Facebook messenger app for the same. The user will send the message to bot. He/she can also send it in the form of image and location.As explained in the architecture, the bot will pass the message to Heroku app and from there will do an API call on diagflow where the actual NLP takes and the response is generated and send back to user.

Activity Diagram

It shows the activity diagram when the user is chatting with the bot. The very first condition is to check whether the logged in or not. If not, the user has to login first. If yes, the user will have navigate to VISA bot page and open the window. The user will send the message in text format to the bot. The will then pass the message to Heroku app and then to diagflow where NLP occurs. At diagflow, the condition is checked if the message is proper or not. If not, the bot will send call back response saying that it didn’t understood the message and ask the user to type again. If message is fine, then the intents is extracted where the keywords are stored and synonyms and response is sent back to user.