DailyArt: Artwork Recommendation Application based on the emotional state
Ahyoung Han, South Korea , Kyung Hoon Hyun, South Korea
For art galleries, it is important to provide sublime artistic experiences to the visiting audiences. However it is difficult to satisfy great number of audiences since everyone has different artistic preferences. In addition, creating revenues by selling the art works is equally as important. Therefore, it is critical to understand the artistic preferences for art galleries to increase their chances of sales. Most applications – mobile application and web services – for art galleries concentrate on providing information on art exhibitions such as location and lists of artists for promotion purposes. The absence of a customized art recommendation service opens up new opportunity for investigations on finding effective communication between art galleries and individual audiences. Thus, this paper suggests system, which allows art galleries to collect audiences’ current emotional states, and recommend artworks that correctly represents the emotional states of individual audiences.
Currently, art galleries stores mailing lists of VIP customers – customers with loyalty and potential customers – for periodic advertisements and promotions to provide information on latest art works, news and exhibition to the target customers. This unilateral way of communication does not consider target audiences’ individual artistic preference and it is critical to find ways of analyzing audiences’ current emotional states to understand the target audiences’ individual artistic preference. Artistic preferences differ based on their current emotional state; the artistic preferences is closely related to the level of sympathy, which varies depending on how depressed or happy they are.
We designed a mobile application to create effective recommendation system. To do that:
1) We analyzed emotional levels of artworks was tested at UNC gallery – South Korean art gallery – by selecting the description of the work created by professional curators through text analysis to a