Embodying Latent Requirements with Unexperienced Attractions through Selection of Travel Point Photographs

摘要

People who want to travel search for travel points satisfying their travel requirements. They search for travel points such that they feel attractions to fulfill the requirements. They want to search for them easily and quickly. In recent years, there are many people who use the Internet to collect information on travel points. They use the Internet because they can collect the information on the travel point anytime and anywhere. On the other hand, there are various kinds of information on a lot of travel points on the Internet. Users have to embody conditions to specify the travel point they want to visit. True pleasures of travels are to experience what they have never experienced. Requirements for travel points to seem to be attractive are quite unclear. For the recommendation of travel points, people want a tool to embody easily and quickly the conditions specifying the travel point they want to visit. Current search engines for travel points on the Internet utilize the user based collaborative filtering from travel histories of travelers. However, they cannot grasp what latent requirements travelers have and what kinds of attraction they expect, because they analyze travel histories. The paper proposes a recommendation method which enables users to embody each potential requirement for their travel beforehand. The method utilizes photographs of travel points to infer what latent travel requirement users have and what kinds of attractions they expect to fulfill the latent requirements. An experiment has proved that almost two thirds of subjects are satisfied with travel points recommended using this method. Furthermore, the method allows all of the subjects to decide on travel points they want to visit in a shorter time than the method they usually use.