KIT405 Assignment 3 Developer Manual PDF

Preview

Summary

KIT405 Assignment 3 Application Developer Manual...

Description

DEVELOPER MANUAL

KIT405 ASSIGNMENT 3

CHATBOT WEB APPLICATION - DEVELOPER MANUAL

Table of Contents A. Allocated Topic ............................................................................................ 2 B. Web Services (APIs) ..................................................................................... 2 1. Geolocation API ........................................................................................ 2 2. Place API ................................................................................................... 2 3. Web Crawling ........................................................................................... 2 C. Dialog Scripts ............................................................................................... 3 D. AI Techniques .............................................................................................. 4 1. NLP Technique: ......................................................................................... 4 2. Opinion Mining & Sentiment Analysis: ..................................................... 4 3. Data Classification: ................................................................................... 4 4. Recommendation with User Data: ............................................................ 5 E. User's Data .................................................................................................. 5 F. Relevant Issues ............................................................................................ 5 G. Screen Recording ......................................................................................... 5 H. Login Credentials ......................................................................................... 5 I.

Acknowledgements ..................................................................................... 5

1

CHATBOT WEB APPLICATION - DEVELOPER MANUAL

A. Allocated Topic

•

•

•

From the Project proposal in Assignment 1 and presentation in Assignment, the topic proposed is building a Chatbot Web Application for tourism that help to recommend users for their hotels, restaurants, and other tourism information for user's holiday planning experiences. The Chatbot application provides a text-to-speech interface using the Web Speech API; the system also combines with other Web Services APIs and AI techniques to provide users with helpful information and recommendations. The main file is chatbot.php

B. Web Services (APIs) 1. Geolocation API – uses to locate user's current location. 2. Place API – uses to recommend user's searches for hotels and restaurants locations. 3. Web Crawling - According to the Urban List website, web Crawling uses this API to crawl for the best hotels in Hobart. https://www.theurbanlist.com/alist/best-hotels-hobart.

2

CHATBOT WEB APPLICATION - DEVELOPER MANUAL

C. Dialog Scripts System User System User System User System User System User System User System User System User System User System User System User System User System User System User System User System User System User System

Hello! Welcome to Chatbot Web Application. Hi/Hello Hello, Hai! It is great to see you here How are you? I am doing great and hope the same for you. Do you need my help? How can I use the system?/Help You can ask me for information about hotels, accommodations, and restaurants. I want to plan for my holiday. Great! I can help you with that. Can I get your current location? Please go ahead. I am checking your location, give me a sec. I am looking for an accommodation Ok! I surely can help you with that, but I will need a few details from you, please fill out the form below. User to fill out the form (budget = 170 & people = 5) We recommend you book for an APARTMENT; they are showing below: I am looking for a hotel instead. The nearby hotels are shown below. Can you recommend me any of them? Hai, according to your previous check-ins, here are the recommended hotels: …

Can you search for the best one? Hai, according to The UrbanList's Best Hotels in Hobart and here are the top 5: …

Wow it is amazing. Thank you for your kind words Hai . Are there anything else I can help you with? Are there any nearby restaurants? The nearby restaurants are shown below. They are bad I do not like it. Sorry that you feel that way Hai . Are there anything else I can help you with?

Can you recommend me a best place to have for dinner? Hai, based on your references, here are the recommended restaurants: … Great! thank you. Thank you for your kind words Hai . Are there anything else I can help you with. No, thank you. It is my pleasure to help you, have a great day Hai Bye Goodbye! See you soon, Hai 3

CHATBOT WEB APPLICATION - DEVELOPER MANUAL

D. AI Techniques 1. NLP Technique: • • •

This AI technique detects specific phrases that use mention during the conversation then execute the function accordingly. When the user mentions 'accommodation', the system recognises that they want to look for a place. Other phrases such as hotel, restaurants, etc. also use to detect the user's speech.

2. Opinion Mining & Sentiment Analysis: • •

•

It uses the NLP technique to catchwords and phrases that indicate the user's mood toward using the system. It is passing the phrases to sentiment analysis via PHP then running the function according to the result. If it is positive, then the system will thank the user and apologise if it is negative. The system will continue to ask the user for the next operation after sentiment analysis.

3. Data Classification: • • •

Using the training data to predict for user's decision when looking for accommodation. The training and testing files are included. Using the Decision Tree model in RapidMiner to perform data mining. Predicting user's decision from the budget and number of people.

4

CHATBOT WEB APPLICATION - DEVELOPER MANUAL

4. Recommendation with User Data: • •

Using the list of ratings for hotels and restaurants by the user to provide the recommendations based on their previous preferences. The PHP item-based filtering is used to analyse the similarities between liking references from one user to the other to draw recommendations.

E. User's Data The following files that included in the Assignment: •

•

•

Training and testing data set for Data Mining: o accomodation-testing.csv o accomodation-training.csv Recommendation with User Data: o list_hotel.php o list_restaurant.php Web Crawler: o web_crawler.php o [https://www.theurbanlist.com/a-list/best-hotels-hobart]

F. Relevant Issues • • • •

The chatbot may not be able to detect noun phrases in few rare instances. Place APIs also not running correctly but fully function after restarting. The system may have trouble detecting the most accurate current location. Other than that, the chatbot is expected to operate normally.

G. Screen Recording •

UTAS OneDrive:

https://universitytasmaniamy.sharepoint.com/:v:/r/personal/lenh_utas_edu_au/Documents/MITS%20Sem%201%2 021/KIT405/Assignment%203/kit405_at3_recording_553491.mp4?csf=1&web=1

H. Login Credentials All API Keys have included in the file and implemented correctly; there is no use of other Web services that required login credentials.

I. Acknowledgements •

This Assignment contains some of the source codes and materials from the tutorial works. 5...