Photo Classification Using Machine Learning to Understand the Interests of Tourists

辻岡 卓
渡辺 公次郎
塚本 章宏

Information Systems for Intelligent Systems. Smart Innovation, Systems and Technologies

Springer, Singapore

https://web.db.tokushima-u.ac.jp/cgi-bin/edb_browse?EID=394723

In this study, we attempt to cluster photos posted on Tripadvisor. Many tourists post photos on travel service platforms, such as Tripadvisor, and these photos represent the interests of the tourists. However, these photos are not often analyzed due to the time and effort involved. With the proposed method, features are extracted, and dimensionality is reduced using a machine learning method. The photos are then automatically classified using an unsupervised learning method. The clustering results showed that similar photos were classified into the same cluster. The proposed method can reduce the cost of interpreting tourists' interests.