ABSTRACT
Nowadays, traveling are known to be a healing activity for those are exhausted or those who seeking for an eye-opening experience. This research focus on traveller; backpackers who always have difficulty in finding a nice and the best accommodation suit them through the existing hotel recommendation systems (RSs). Most backpackers still relying on family and friends recommendation instead of looking for searching for accommodation online. One of the problems face by backpacker is that there are only a few accommodation RSs are hostel based. It may not be a good option for backpacker to search for accommodation in hotel based RSs as the price might be a little expensive for them. Other than that, most existing RSs for backpacker does not produced an accurate result because the criteria taken are normally less than three. Hence the result they get are always based on destination and date chosen only. The design and functionality offered in existing accommodation RSs for backpacker are also confusing and not well developed. To counter the problems face by them, this Backpacker Recommendation System (BRS) is developed based on the survey did with the help of 50 respondent from Backpackers Buddies Malaysia (Facebook). BRS taking in eight criteria from user to produce a list of accommodation suits the backpacker based on his traveling and accommodation preferences. The criteria included are destination, price range, travel types, room types, private locker, food/beverage offer, trip package/advise and pickup service. Moreover, the technique used in BRS is data mining in details; Decision Tree Algorithm. The BRS were designed to assists backpackers in searching for the best available accommodation. This system enhance the accuracy of accommodation RSs with the help of decision tree algorithm.
INTRODUCTION
Backpackers’ Accommodation Recommendation System (BARS) is a system that provides the best possible accommodation suggestion based on the backpacker’s preferences. This system is designed specifically for backpacker and is focusing on the different criteria of the accommodation. The idea of developing this system comes from the weaknesses in existing accommodation RS. The existing systems does not provide personalized recommendation criteria and some proposed inefficient RS for users. This project propose RS based on backpacker’s preferences on lists of criteria before making a suggestion.