imprint:
Association for the Advancement of Artificial Intelligence (AAAI), 2021
Published in:Proceedings of the International AAAI Conference on Web and Social Media
Language:
Not determined
DOI:
10.1609/icwsm.v10i1.14736
ISSN:
2334-0770;
2162-3449
Origination:
Footnote:
Description:
<jats:p>
People are able to detect up to 1 trillion odors. Yet, city planning is concerned only with a few bad odors, mainly because odors are currently captured only through complaints made by urban dwellers. To capture both good and bad odors, we resort to a methodology that has been recently proposed and relies on tagging information of geo-referenced pictures. In doing so for the cities of London and Barcelona, this work makes three new contributions. We study 1) how the urban smellscape changes in time and space; 2) which emotions people share at places with specific smells; and 3) what is the color of a smell, if it exists. Without social media data, insights about those three aspects have been difficult to produce in the past, further delaying the creation of urban restorative experiences.
</jats:p>