[H-TT19] Geographic Information System and Cartography

[HTT19-07] On the seasonal colors and smells analyzing Social Networking Service

*Tatsuya MATSUURA¹, Kazunari TANAKA² (1.Graduate School of Engineering Osaka Institute of Technology, 2.Osaka Institute of Technology)

Keywords:Colors, Smells, Twenty-four Divisions of the Solar Year, SNS

The purpose of this study is to examine the method to extract the sense of the season by colors and smells.The research method is based on the Twenty-four Divisions of the Solar Year as a basic unit, extract colors on the photos and smells on the text on SNS ; Twitter and Facebook, which are people feel the season. We collect the data using Geospatial Information System and text mining software. First, we collect 2017 yearly tweets posted on Twitter. The photo containing the tweet is subjected to extract the color of the season.Sentences are subjected to text mining to extract smells. In this research, we try to extract not typical few colors of every seasons and smells but all colors and smells with their relationships. The result shows that there are from Sixty four to Seventy seven colors as seasonal colors. The smells of the season is the first half of February, the second half of May, the second half of June, the first half of July, the second half of September, October. Other seasons can’t feel smells of the season.