Can supervise: YES
Zhang, Q, Lu, J, Wu, D & Zhang, G 2019, 'A Cross-Domain Recommender System With Kernel-Induced Knowledge Transfer for Overlapping Entities.', IEEE transactions on neural networks and learning systems, vol. 30, no. 7, pp. 1998-2012.View/Download from: UTS OPUS or Publisher's site
The aim of recommender systems is to automatically identify user preferences within collected data, then use those preferences to make recommendations that help with decisions. However, recommender systems suffer from data sparsity problem, which is particularly prevalent in newly launched systems that have not yet had enough time to amass sufficient data. As a solution, cross-domain recommender systems transfer knowledge from a source domain with relatively rich data to assist recommendations in the target domain. These systems usually assume that the entities either fully overlap or do not overlap at all. In practice, it is more common for the entities in the two domains to partially overlap. Moreover, overlapping entities may have different expressions in each domain. Neglecting these two issues reduces prediction accuracy of cross-domain recommender systems in the target domain. To fully exploit partially overlapping entities and improve the accuracy of predictions, this paper presents a cross-domain recommender system based on kernel-induced knowledge transfer, called KerKT. Domain adaptation is used to adjust the feature spaces of overlapping entities, while diffusion kernel completion is used to correlate the non-overlapping entities between the two domains. With this approach, knowledge is effectively transferred through the overlapping entities, thus alleviating data sparsity issues. Experiments conducted on four data sets, each with three sparsity ratios, show that KerKT has 1.13%-20% better prediction accuracy compared with six benchmarks. In addition, the results indicate that transferring knowledge from the source domain to the target domain is both possible and beneficial with even small overlaps.
Zhang, Q, Wu, D, Lu, J, Liu, F & Zhang, G 2017, 'A cross-domain recommender system with consistent information transfer', Decision Support Systems, vol. 104, pp. 49-63.View/Download from: UTS OPUS or Publisher's site
© 2017 Elsevier B.V. Recommender systems provide users with personalized online product and service recommendations and are a ubiquitous part of today's online entertainment smorgasbord. However, many suffer from cold-start problems due to a lack of sufficient preference data, and this is hindering their development. Cross-domain recommender systems have been proposed as one possible solution. These systems transfer knowledge from one domain that has adequate preference information to another domain that does not. The outlook for cross-domain recommendation is promising, but existing methods cannot ensure the knowledge extracted from the source domain is consistent with the target domain, which may impact the accuracy of the recommendations. To address this challenging issue, we propose a cross-domain recommender system with consistent information transfer (CIT). Knowledge consistency is based on user and item latent groups, and domain adaptation techniques are used to map and adjust these groups in both domains to maintain consistency during the transfer learning process. Experiments were conducted on five real-world datasets in three categories: movies, books, and music. The results for nine cross-domain recommendation tasks show that CIT outperforms five benchmarks and increases the accuracy of recommendations in the target domain, especially with sparse data. Practically, our proposed method is applied into a telecom product recommender system and a business partner recommender system (Smart BizSeeker) to enhance personalized decision making for both businesses and individual customers.
Zhang, Q, Lu, J, Wu, D & Zhang, G 2018, 'Cross-domain recommendation with consistent knowledge transfer by subspace alignment', Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), International Conference on Web Information Systems Engineering, Springer Link, Dubai, United Arab Emirates, pp. 67-82.View/Download from: UTS OPUS or Publisher's site
© Springer Nature Switzerland AG 2018. Recommender systems have drawn great attention from both academic area and practical websites. One challenging and common problem in many recommendation methods is data sparsity, due to the limited number of observed user interaction with the products/services. Cross-domain recommender systems are developed to tackle this problem through transferring knowledge from a source domain with relatively abundant data to the target domain with scarce data. Existing cross-domain recommendation methods assume that similar user groups have similar tastes on similar item groups but ignore the divergence between the source and target domains, resulting in decrease in accuracy. In this paper, we propose a cross-domain recommendation method transferring consistent group-level knowledge through aligning the source subspace with the target one. Through subspace alignment, the discrepancy caused by the domain-shift is reduced and the knowledge shared local top-n recommendation via refined item-user bi-clustering two domains is ensured to be consistent. Experiments are conducted on five real-world datasets in three categories: movies, books and music. The results for nine cross-domain recommendation tasks show that our proposed method has improved the accuracy compared with five benchmarks.
Zhang, Q, Wu, D, Lu, J & Zhang, G 2018, 'Cross-domain recommendation with probabilistic knowledge transfer', Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), pp. 208-219.View/Download from: UTS OPUS or Publisher's site
© Springer Nature Switzerland AG 2018. Recommender systems have drawn great attention from both academic and practical area. One challenging and common problem in many recommendation methods is data sparsity, due to the limited number of observed user interaction with the products/services. To alleviate the data sparsity problem, cross-domain recommendation methods are developed to share group-level knowledge in several domains so that recommendation in the domain with scarce data can benefit from domains with relatively abundant data. However, divergence exists in the data of similar domains so that the extracted group-level knowledge is not always suitable to be applied in the target domain, thus recommendation accuracy in the target domain is impaired. In this paper, we propose a cross-domain recommendation method with probabilistic knowledge transfer. The proposed method maintain two sets of group-level knowledge, profiling both domain-shared and domain-specific characteristics of the data. In this way users’ mixed preferences can be profiled comprehensively thus improves the performance of the cross-domain recommender systems. Experiments are conducted on five real-world datasets in three categories: movies, books and music. The results for nine cross-domain recommendation tasks show that our proposed method has improved the accuracy compared with five benchmarks.
Zhang, Q, Wu, D, Zhang, G & Lu, J 2016, 'Fuzzy user-interest drift detection based recommender systems', 2016 IEEE International Conference on Fuzzy Systems, FUZZ-IEEE 2016, IEEE International Conference on Fuzzy Systems, IEEE, Vancouver, Canada, pp. 1274-1281.View/Download from: UTS OPUS or Publisher's site
© 2016 IEEE.Recommender systems aim to provide personalized suggestions to users by modeling user-interests to deal with information overload problem, which is extremely severe in the era of big data. Since user-interests are drifting due to their taste variation on items, recommender systems without considering that will suffer degradation of prediction accuracy. There are two challenges about adapting to user-interest drift in recommender systems: 1) accurately modeling user-interests is not easy since the drift of user-interests may occur in different direction for each user; 2) item features and user-interests are often incomplete and vague, which makes it more difficult to model user-interests. To handle these two issues, this study proposes a fuzzy user-interest drift detection based recommender system that adapts to user-interest drift and improves prediction accuracy. A fuzzy user-interest consistency model is built based on fuzzy set theories, and a user-interest drift detection approach and algorithms are developed based on concept drift techniques to provide guidance to recommendation generation. Empirical experiments are conducted on synthetic and real-world MovieLens datasets. The results show that the proposed approach improves the performance of recommender systems in metric of MAE.
Zhang, Q, Zhang, G, Lu, J & Wu, D 2015, 'A framework of hybrid recommender system for personalized clinical prescription', Proceedings of the 10th International Conference on Intelligent Systems and Knowledge Engineering (ISKE 2015), International Conference on Intelligent Systems and Knowledge Engineering, IEEE, Taipei Taiwan, pp. 189-195.View/Download from: UTS OPUS or Publisher's site
General practitioners are faced with a great challenge of clinical prescription owing to the increase of new drugs and their complex functions to different diseases. A personalized recommender system can help practitioners deal with mass of medical knowledge hidden in history medical records. To support practitioner’s decision making in prescription, this paper proposes a framework of a hybrid recommender system which integrates artificial neural network and case-based reasoning. Three issues are considered in this system framework: (1) to define a patient’s need by giving his/her symptom, (2) to mine features from free text in medical records and (3) to analyze temporal efficiency of drugs. The proposed recommender system is expected to help general practitioners to improve their efficiency and reduce risks of making errors in daily clinical consultation with patients.