3.1 Tourism and technology
Tourism is traveling, mainly for recreational purposes or utilizing the leisure or to know and experience different cultures and heritage. Tourism is the fastest grow- ing industry with a huge potential to generate employment for people and rev- enue for the government. Emergence of Information and Communication Tech- nology (ICT) and internet transformed the way business is done worldwide. Tourism industry is one of the early adapter of technology and internet. Ever since internet emerged, people predominantly used internet for travel planning (e.g., travel information search and booking). With internet penetration growing more and more, importance of technology in tourism is also growing more and more. Technology and internet have been rapidly applied and diffused through tourism sectors. This is where E-Tourism comes into picture. Technology is one of the external environment elements for tourism, travel and hospitality. It offers the bridge of communication between consumers and suppliers.
Technology changes how travelers access and use travel-related information. Many businesses and locations start the experience before a visitor arrives. Pre-travel preparation is an important phase in the process because it allows the customer to learn more about his upcoming trip. The internet is one of the most useful tools for trip planning. The tour guide can now be a GPS tour guide, the guidebook can be an audio guide, and trips can be scheduled entirely online. The advancement of information technology, as well as widespread public use of the Internet, has created a slew of conditions that have impacted the modern travel agency in both positive and negative ways. Many areas of industry are being altered by the internet. As a result, in the future, the travel and tourism industries will have to continue to adapt to emerging technologies[4].
3.2 Convolutional Neural Network
Convolutional Neural Network (CNN or ConvNet) is a form of multi-layer neural network that is driven by the optical system of human beings. A typical CNN is made up of one or more completely connected layers, followed by one or more blocks of convolution and subsampling layers, and finally an output layer. CNN is a Deep Learning algorithm that can take an image as input, assign learnable weights and prejudices to various features/objects in the image, and recognise each image as unique. In comparison to other classification algorithms, ConvNet needs significantly less pre-processing. ConvNets can learn these filters/characteristics with enough experience, while primitive methods require hand-engineering.
The key component of a CNN is the convolutional layer (conv layer). In nature, images are usually static. That is, the forming of one part of the image is identical to the formation of every other part. As a result, a function learned in one area may be used to fit a pattern in another. We take a small part of a large image and transfer it through all of the points in the large image (Input). We condense them into a single location when going through some point (Output). Filter (Kernel) refers to each small portion of the image that passes over the larger image. CNN is a layered network that is completely linked. This layer takes input from all neurons in the previous layer and generates output by performing operations on individual neurons in the current layer.
3.3 Google Map API
You can use the Google Maps Android API to include maps and interactive mapping data in your app. We can add maps to your application that are based on Google Maps data. Using the Maps SDK for Android, the API automatically handles access to Google Maps servers, data download, map view, and response to map gestures. API calls enable you to customize a simple map by adding markers, polygons, and overlays, as well as changing the user's view of a specific map region. These objects give users more detail about map positions and enable them to interact with the map. It helps users to discover locations using Google's detailed maps. Custom markers can be used to identify positions, image overlays can be used to supplement map data, one or more maps can be embedded as fragments, and so much more.
3.4 Collaborative Filtering
When it comes to developing intelligent recommender systems that can learn to provide better recommendations as more knowledge about users is collected, collaborative filtering is the most commonly used technique.
Collaborative filtering is a method of distinguishing objects that a user would like based on the reactions and feedback of other users. It works by sifting through a large number of people to find a smaller group of people who have similar tastes to a specific person. It analyses their favorite products and compiles a ranked list of recommendations. There are several methods for determining which users are similar and combining their choices to generate a list of suggestions.
To develop a framework that can automatically suggest things to users based on the interests of other users. The first move is to look for users or artefacts that are similar. Estimating ratings of objects that have not yet been rated by a consumer is the second stage.