IRLGAug 8, 2023

RECipe: Does a Multi-Modal Recipe Knowledge Graph Fit a Multi-Purpose Recommendation System?

arXiv:2308.04579v11 citationsh-index: 11
Originality Incremental advance
AI Analysis

This addresses recipe recommendation for users on online platforms, offering a novel multi-modal approach but is incremental in combining existing techniques.

The paper tackles recipe recommendation by introducing RECipe, a multi-purpose framework using a multi-modal knowledge graph with subsystems for behavior-, review-, and image-based recommendations, achieving performance comparable to neural solutions on public datasets.

Over the past two decades, recommendation systems (RSs) have used machine learning (ML) solutions to recommend items, e.g., movies, books, and restaurants, to clients of a business or an online platform. Recipe recommendation, however, has not yet received much attention compared to those applications. We introduce RECipe as a multi-purpose recipe recommendation framework with a multi-modal knowledge graph (MMKG) backbone. The motivation behind RECipe is to go beyond (deep) neural collaborative filtering (NCF) by recommending recipes to users when they query in natural language or by providing an image. RECipe consists of 3 subsystems: (1) behavior-based recommender, (2) review-based recommender, and (3) image-based recommender. Each subsystem relies on the embedding representations of entities and relations in the graph. We first obtain (pre-trained) embedding representations of textual entities, such as reviews or ingredients, from a fine-tuned model of Microsoft's MPNet. We initialize the weights of the entities with these embeddings to train our knowledge graph embedding (KGE) model. For the visual component, i.e., recipe images, we develop a KGE-Guided variational autoencoder (KG-VAE) to learn the distribution of images and their latent representations. Once KGE and KG-VAE models are fully trained, we use them as a multi-purpose recommendation framework. For benchmarking, we created two knowledge graphs (KGs) from public datasets on Kaggle for recipe recommendation. Our experiments show that the KGE models have comparable performance to the neural solutions. We also present pre-trained NLP embeddings to address important applications such as zero-shot inference for new users (or the cold start problem) and conditional recommendation with respect to recipe categories. We eventually demonstrate the application of RECipe in a multi-purpose recommendation setting.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes