A Unified Collaborative Representation Learning for Neural-Network Based Recommender Systems

Abstract

With the boosting of neural networks, recommendation methods become significantly improved by their powerful ability of prediction and inference. Existing neural-network based recommender systems (NN-RSs) usually first employ matrix embedding (ME) as a pre-process to learn users' and items' representations (latent vectors), then input these representations to a specific modified neural network framework to make accurate Top-k recommendations. Obviously, the performance of ME has a significant effect on RS models. However, most NN-RSs focus on accuracy by building representations from the direct user-item interactions (e.g., user-item rating matrix), while ignoring the underlying relatedness between users and items (e.g., users who rate the same ratings for the same items should be embedded into similar representations), which is an ideological disadvantage. On the other hand, ME models directly employ inner products as a default loss function metric that cannot project users and items into a proper latent space, which is a methodological disadvantage. In this paper, we propose a supervised collaborative representation learning model -Magnetic Metric Learning (MML) -to map users and items into a unified latent vector space, enhancing the representation learning for NN-RSs. Firstly, MML utilizes dual triplets to model not only the observed relationships between users and items, but also the underlying relationships between users as well as items to overcome the ideological disadvantage. Specifically, a modified metric-based dual loss function is proposed in MML to gather similar entities and disperse the dissimilar ones. With MML, we can easily compare all the relationships (user to user, item to item, user to item) according to the weighted metric, which overcomes the methodological disadvantage. We conduct extensive experiments on four real-world datasets with large item space. The results demonstrate that MML can learn a proper unified latent space for representations from the user-item matrix with high accuracy and effectiveness, and lead to a performance gain over the state-of-the-art RS models by an average of 17%.

ODTRA-Based Task Offload Optimization for Industrial Internet of Things: Improving Efficiency and Performance With Digital Twins and Metaheuristic Optimization

“…They explored several sectors, such as healthcare, industry, and smart cities. They highlighted technologies used in creating DT, including IoT, IIoT, REST, SOAP, cloud computing, Machine Learning (ML): (a) Supervised Learning (SL) and (b) Unsupervised Learning (UL), and Deep Learning (DL) [52]), diverse databases (MongoDB, Redis, MySQLi), and data analytics with recognized smart manufacturing as an advanced model with cognitive abilities, proposing a DT-based model that employed intelligent methods for autonomous manufacturing [53]. [54] discussed the emergence of Industry 4.0, which promotes modernizing traditional manufacturing through technology-driven approaches.…”

Human Origin‐Destination Flow Prediction Based on Large Scale Mobile Signal Data

“…In recent years, various machine learning and deep learning models have been widely used in traffic OD flow prediction. These models represent the complex nonlinear relationship between different variables and provide a new prediction method [12,13].…”

Adaptive habitat biogeography-based optimizer for optimizing deep CNN hyperparameters in image classification

“…These parameters are referred to as DCNN’s hyperparameters. The hyperparameters' tuning referred to the task of determining the right values of hyperparameters for a particular problem [ 35 ]. The optimization of hyperparameters is an Nondeterministic Polynomial (NP)-hard problem, which is one of the primary challenges confronting DCNN’s training [ 36 ].…”