Emotional Insights for Food Recommendations
Rostami, Mehrdad; Vardasbi, Ali; Aliannejadi, Mohammad; Oussalah, Mourad (2024-03-16)
Springer
16.03.2024
Rostami, M., Vardasbi, A., Aliannejadi, M., Oussalah, M. (2024). Emotional Insights for Food Recommendations. In: Goharian, N., et al. Advances in Information Retrieval. ECIR 2024. Lecture Notes in Computer Science, vol 14609. Springer, Cham. https://doi.org/10.1007/978-3-031-56060-6_16
https://rightsstatements.org/vocab/InC/1.0/
© 2024 The Author(s), under exclusive license to Springer Nature Switzerland AG
https://rightsstatements.org/vocab/InC/1.0/
© 2024 The Author(s), under exclusive license to Springer Nature Switzerland AG
https://rightsstatements.org/vocab/InC/1.0/
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:oulu-202404182839
https://urn.fi/URN:NBN:fi:oulu-202404182839
Tiivistelmä
Abstract
Food recommendation systems have become pivotal in offering personalized suggestions, enabling users to discover recipes in line with their tastes. However, despite the existence of numerous such systems, there are still unresolved challenges. Much of the previous research predominantly lies on users’ past preferences, neglecting the significant aspect of discerning users’ emotional insights. Our framework aims to bridge this gap by pioneering emotion-aware food recommendation. The study strives for enhanced accuracy by delivering recommendations tailored to a broad spectrum of emotional and dietary behaviors. Uniquely, we introduce five novel scores for Influencer-Followers, Visual Motivation, Adventurous, Health and Niche to gauge a user’s inclination toward specific emotional insights. Subsequently, these indices are used to re-rank the preliminary recommendation, placing a heightened focus on the user’s emotional disposition. Experimental results on a real-world food social network dataset reveal that our system outperforms alternative emotion-unaware recommender systems, yielding an average performance boost of roughly 6%. Furthermore, the results reveal a rise of over 30% in accuracy metrics for some users exhibiting particular emotional insights.
Food recommendation systems have become pivotal in offering personalized suggestions, enabling users to discover recipes in line with their tastes. However, despite the existence of numerous such systems, there are still unresolved challenges. Much of the previous research predominantly lies on users’ past preferences, neglecting the significant aspect of discerning users’ emotional insights. Our framework aims to bridge this gap by pioneering emotion-aware food recommendation. The study strives for enhanced accuracy by delivering recommendations tailored to a broad spectrum of emotional and dietary behaviors. Uniquely, we introduce five novel scores for Influencer-Followers, Visual Motivation, Adventurous, Health and Niche to gauge a user’s inclination toward specific emotional insights. Subsequently, these indices are used to re-rank the preliminary recommendation, placing a heightened focus on the user’s emotional disposition. Experimental results on a real-world food social network dataset reveal that our system outperforms alternative emotion-unaware recommender systems, yielding an average performance boost of roughly 6%. Furthermore, the results reveal a rise of over 30% in accuracy metrics for some users exhibiting particular emotional insights.
Kokoelmat
- Avoin saatavuus [37782]