Content-based artwork recommendation: integrating painting metadata with neural and manually-engineered visual features

Trends in content-based recommendation

“…New types of content Some domains are rich in non-textual content, which is currently still difficult to utilize effectively, such as multimedia content. So far, multimedia recommendation mostly relies on implicit and explicit user feedback, but we expect that the rapid developments in DL algorithms will enable more structured and effective ways of extracting and using semantic and stylistic features in the future, as done in Messina et al (2018) in this issue. Open research questions here include how and which features can be extracted from multimedia content, and how they should be integrated with user-generated content and user preference data to provide the most relevant recommendations.…”

“…The paper Content-Based Artwork Recommendation: Integrating Painting Metadata with Neural and Manually-Engineered Visual Feature (Messina et al 2018) introduces a novel approach for artwork recommendation. The authors combine a variety of information sources-including item metadata as well as low-level visual features that are extracted through a deep neural network-and show the advantage of their combined approach based on an experiment with real-world data provided by an online artwork store.…”

Visual Cultural Biases in Food Classification

“…The ten features in this set represent low-level image properties including image brightness, sharpness, contrast, colorfulness, entropy, RGB (Red Green Blue) contrast, variation in sharpness, saturation, variation in saturation and naturalness. These features are simple to calculate and have shown utility in several image popularity predictions and recommendation tasks, from the photos in Folksonomies [ 16 ] to specific categories of images, such as recipe images [ 15 ] and artwork [ 17 ]. The freely available OpenIMAJ ( ) framework was employed to calculate the EVF features.…”

Juger du Beau avec subjectivité : le défi de l’esthétique computationnelle

“…Dans le domaine de la vente d'oeuvres d'art en ligne, des services de recommandation existent également [Dominguez et al, 2017, Benouaret, 2017, Messina et al, 2018, He et al, 2016, qui semblent s'approcher de notre problème. Dans ces systèmes, il apparaît cependant que, si la peinture ou la photo entrent bien en ligne de compte dans la recommandation par leur style et leur composition, le rôle de ces informations issues de l'image reste mineur face au rôle considérable de données plus sémantiques comme le nom de l'auteur, sa cote dans le marché, le prix de l'oeuvre, ou les achats antérieurs du client.…”