Construction and demonstration experiment of similar spot recommendation system based on emotion word contained in word-of-mouth information

Abstract

In order to increase the staying time in sightseeing in local cities, this study calculates a spot similar to the emotion score of the selected spot, and based on the emotion word contained in the word of mouth information that makes the spot with a smaller difference similar Build a similar spot recommendation system. This system collects posted word-of-mouth information, calculates a score by emotion analysis, and then extracts highly attractive spots. Next, demonstration experiment is done in Sakai City, Osaka in a study cooperation destination using the main system, and the performance value about the system is evaluated. Furthermore, for the spot recommendation function of the two main functions of this system, the ratio of accurate judgment based on the error between the subjective evaluation and the emotion analysis by natural language processing is clarified.