A Survey of Recommender Systems with Multi-Objective Optimization

A survey of recommender systems with multi-objective optimization

Medicine recommendation systems target to recommend a set of medicines given a set of symptoms which play a crucial role in assisting doctors in their daily clinics. Existing approaches are either rule-based or supervised. However, the former heavily relies on expert labeling, which is time-consuming and costly to collect, and the latter suffers from the data sparse problem. To automate medicine recommendation on sparse data, we propose MedRec, which introduces two graphs in modeling: (1) a knowledge graph connecting diseases, medicines, symptoms, and examinations; (2) an attribute graph connecting medicines via shared attributes and molecular structures. These two graphs enhance the connectivity between symptoms and medicines, which thus alleviate the data sparse problem. By learning the interrelationship between diseases, medicines, symptoms and examinations and the inner relationship within medicine, we can acquire unified embedding representations of symptoms and medicines which can be used in medicine recommendation. The experimental results show that the proposed model outperforms state-of-the-art methods. In addition, we find that these two tasks: learning graph representation and medical recommendation can benefit each other.

Knowledge-Enhanced Attributed Multi-Task Learning for Medicine Recommendation

Aiming at the problems of low data quality and high incentive costs caused by the low enthusiasm of participants in mobile crowd sensing, a new task recommendation framework is proposed in this paper First, the participants' historical behaviors are analyzed, assuming that user behaviors can be quantified as the user's willingness to participate, and the cosine similarity theorem is used to calculate the similarity between participants, thereby constructing a user-hybrid model Secondly, probabilistic matrix factorization is developed to predict the willingness of participants, and a ranking model is obtained through learn-to-rank algorithm Finally, a task recommendation list is generated according to the ranking model, which serves as the target participant's preferred task list for sensing task recommendation The experiment in this paper is carried out on the MATLAB platform based on two real check-in datasets, Gowalla and Brightkite The results show that the average allocation precision rate can reach 96%, and the sensing user participation rate is about 97% Meantime, the user's mobile cost is reduced, and the overall goal of maximizing accuracy and minimizing perceived cost is achieved

A task recommendation framework for heterogeneous mobile crowdsensing

Bilateral knowledge graph enhanced online course recommendation

Collaborative filtering (CF) usually suffers from data sparsity and cold-start problems in real recommendation scenarios, therefore, side information like social networks and contexts have been introduced to improve its performance. In this paper, we consider the knowledge graph (KG) as a source of side information and propose a novel framework, Knowledge-aware representation Graph Convolutional Network for Recommendation (KCRec), that is an end-to-end framework that captures the inter-user and inter-item relatedness effectively. For exploring the potential long-distance interests of the user, we aggregate the item features and get the representation of the user preferences by propagating the relationships in KG between their neighborhood, and further integrates with the graph convolution network. Furthermore, we employ similarity features in different users to construct a user-adjacency graph, and utilize the user–item interaction features to establish a user-feature graph, to obtain the high-order representation of users. Extensive experiments on two real-world datasets demonstrate that our proposed method has substantially improved, which outperforms several state-of-the-art baselines.

KCRec: Knowledge-aware representation Graph Convolutional Network for Recommendation

User-item rating interactions in the recommender system have a deep potential connection with the friend relationships in the social network. In short, users who like the same kind of items may be potential friends in social network, and vice versa, friends in social networks tend to like similar items. Although the above-mentioned two kinds of interactive information can complement and inspire each other, either of them is sparse, which is still not enough to make accurate recommendations. In order to make up for this defect, we then mine useful information from attribute information, learning more informative node representation. In this paper, we explore attribute learning and mutual utilization, complementation and inspiration between social data and rating data. We propose a generic Attribute-Aware Multi-task Recommendation framework (AAMR) for rating prediction and social prediction, which learns representations for users and items by preserving both rating data and social data and attribute information, so as to conduct both rating prediction and trust relationship prediction tasks. Because many users are both in the rating matrix and in social networks, in the common learning, the two tasks will share the embedding of users, which makes the social data and rating data enrich each other’s semantics and alleviate each other’s sparsity. To justify our proposal, we conduct extensive experiments on a real-world dataset. Compared to the state-of-the-art rating and trust prediction approaches, AAMR can learn more informative representations, achieving substantial gains on both tasks.

Attribute-aware multi-task recommendation

Since the rich semantics of attribute information has become a great supplement to the ratings data in designing recommender systems, fusing attributes information into ratings has shown promising performance for many recommendation tasks. However, the use of attribute information is not easy, because different attributes are often: (1) multi-source, that is, attributes may come from many different fields, (2) unstructured, (3) unbalanced, (4) heterogeneous. In this paper, we explore effective fusion of user-item ratings and item attributes to improve recommendations, we propose an attribute-aware deep attentive recommendation model, which embeds attribute information into the latent semantic space of items through the attention mechanism, forming more accurate item representations. Extensive experiments show that our method is superior to the existing methods on both rating prediction and Top-N Recommendation tasks.

Lisa Zhang

Papers

KCRec: Knowledge-aware representation Graph Convolutional Network for Recommendation

Attribute-aware multi-task recommendation

Attribute-aware deep attentive recommendation