Extracting feeling and life-style semantics from hotel images using colour harmony

Abstract:

An application of a cognitively inspired qualitative colour theory, QCharm, which is able to construct harmonic colour palettes based on a model for qualitative colour description (QCD) is presented in this paper. Using Kobayashi's colour space, cognitive keywords representing a feeling or a lifestyle can be assigned to QCharm harmonic-colour palettes. Moreover, using COLOURlovers dataset for learning, the QCharm colour palettes can be rated representing users' preferences. In this paper, QCharm is applied to hotel images (extracted from ImageNet1 image database) as a basis to create a recommender system of hotel marketing materials. The future recommender system is intended to be used to rise the impact of the marketing materials among prospective tourists and then improve the touristic image of the hotel. Furthermore, by adding other cognitive aspects (as a style of life for instance) the system can be used to design customized materials for different types of clients.

Año de publicación:

2019

Keywords:

• Machine learning
• Recommender system
• Cognitive image
• Preferences
• Touristic image
• data science
• tourism marketing
• sentiment analysis
• Colour names

Fuente:

google

scopus