@techreport{oai:ipsj.ixsq.nii.ac.jp:00232667,
 author = {荻原, 幸伽 and 間邊, 哲也 and Sachika, Ogiwara and Tetsuya, Manabe},
 issue = {10},
 month = {Feb},
 note = {本稿では，観光支援に向け，既存のデジタルマップなどで観光スポットとして未登録ではあるものの，観光資源が存在する可能性のあるスポット（以後，「埋もれた観光 POI  (Point of Interest)」）を位置情報ビッグデータから抽出することを目的としている．「埋もれた観光 POI」の抽出に向け，3 種類の抽出方法を考案し，性能を評価した．考案した抽出方法は，既存の方法をアレンジした方法，観測人数の時間変化に注目した方法，観測人数の局所性に注目した方法である．各抽出方法の性能を評価した結果，観測人数の時間変化に注目した方法の抽出性能が最も高かった．次に 2 種類の抽出方法を組み合わせることで抽出性能の向上を図った．この結果，既存の方法をアレンジした方法と観測人数の局所性に注目した方法の組み合わせの抽出性能が最も高くなり，抽出方法を組み合わせることによって相乗効果がもたらされることが分かった．以上のことから，抽出方法の方針に沿った「埋もれた観光 POI」が抽出でき，観光支援に繋がる知見を得ている．, The aim of this paper is to extract “Unidentified sightseeing POI (Point of Interest)” from location-based big data to support sightseeing. “Unidentified sightseeing POI” is the unconfirmed sightseeing spots that are not registered on existing digital maps as sightseeing spots. Three kinds of method devised to extract “Unidentified sightseeing POI”. The methods were devised as follows: the method that improved the conventional research, the method to extract the places where the number of people changed temporarily, and the method to extract the places that are more people than surrounding areas. The performance of each extracting method was evaluated. As a result, the extracting method of the highest performance was the method to extract the places where the number of people changed temporarily. We aimed to enhance performance by combining two kinds of methods. As a result, the combination that became highest in extraction performance was as follows: the method that improved the conventional research and the method to extract the places that are more people than surrounding areas. Both methods did not perform well individually, but when used together, they had a synergistic effect. Consequently, according to the extraction policy, we were able to extract “Unidentified sightseeing POI”. We obtained useful knowledge for sightseeing support.},
 title = {観光支援に向けた埋もれた観光POI抽出に関する一検討},
 year = {2024}
}