Point of Interest Description
Points of Interest in Destination
A recommendation based travel web application that recommends you places to visit and helps plan journeys. The whole process of planning a trip to choosing places to visit, to checking flight proces is a much of a hassle. Travelytics is a solution in that direction that intends to make travel easier. Travelytics recommends and suggests places that you can visit based on your friends circles travel/check-in activity. Check-in activity data of your friends is extracted from Facebook and collaborative filtering is performed to get the top places that you might like to visit. These places are then ranked by getting the top travel blogs for these places from the web by using the Microsoft Cognitive Services, Bing Search API, and performing sentiment analysis on these pages to measure the popularity of the locations by using the IBM Watson Alchemy API. Once a location has been chose, the application helps you plan the rest of you itinerary. The Amadeus Travel Innovation Sandbox API catalogue is used to find the best flight deal and also points of interest at the destination location. There are a lot of API services that are used in this application and are all hosted on Cycle, which hosts all the containers. So all in all it helps you plan your journey more seamlessly and gives you a hassle free experience.
Cycle.io which is a container deployment tool, was used for deploying the containers we needed for our project.
API's Used: Amadeus APIs - Nearest airports of feasible flights, flight discovery, prices, finding nearest places. Microsoft Cognitive Services - Bing search api to extract blogs related to places. Facebook API - Login and get friends and checkin data. IBM Watson API - Data extraction from URLs and sentiment analysis on them. Cycle Infrastructure - Hosting 3 of 4 micro services on docker and cycle.io Google Maps API
Cycle Containers IP Addresses : 184.108.40.206:5000 - Amadeus API 220.127.116.11:5000 - Microsoft Bing Search API 18.104.22.168:5000 - Watson Recommendation API