Scholarlink

Recently, online recommendation systems that recommend travel products or services have been widely used. Many companies are working to develop effective recommendation systems to improve customer interaction and gain competitiveness. This paper is a study on the development of a tourist attraction recommendation system in Jeju Island using embedding based on the Word2Vec algorithm. In this study, we propose a method to predict the place to be visited next by considering the places the traveler has visited as words and the entire itinerary as a sentence. We propose a study that recommends nearby tourist destinations by applying the Word2Vec algorithm to tourists’ travel data and then learning the pattern by Xgboost Classifier model. Based on the results predicted through a machine learning model, nearby vectors are found and the five optimal tourist destinations are recommended. It was confirmed that the proposed method has higher accuracy when the mouth bedding vector has a 4-dimensional dimension than when it is 1-dimensional.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Tourist Attraction Recommendation Using a Word2Vec-Based Approach Enhanced by XGBoost Classifier

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Jiyoung Ko,
  • Jisoo Kim,
  • Yung-Cheol Byun

摘要

Recently, online recommendation systems that recommend travel products or services have been widely used. Many companies are working to develop effective recommendation systems to improve customer interaction and gain competitiveness. This paper is a study on the development of a tourist attraction recommendation system in Jeju Island using embedding based on the Word2Vec algorithm. In this study, we propose a method to predict the place to be visited next by considering the places the traveler has visited as words and the entire itinerary as a sentence. We propose a study that recommends nearby tourist destinations by applying the Word2Vec algorithm to tourists’ travel data and then learning the pattern by Xgboost Classifier model. Based on the results predicted through a machine learning model, nearby vectors are found and the five optimal tourist destinations are recommended. It was confirmed that the proposed method has higher accuracy when the mouth bedding vector has a 4-dimensional dimension than when it is 1-dimensional.