Narrow Results

Aiming to recommend potential collaborators for academic entities such as researchers and institutions, this paper develops a social recommender system through bibliometric indicators and network analytics. Targeting to scholarly articles, the proposed recommender system exploits co-authorships as established social relations and proposes a link prediction model for discovering such potential relations in terms of a co-authorship network. A case study recommending scientific collaborators for research entities on generelated diseases demonstrates the reliability of this study.

In the fast pace of life, E-learning has become a new way for self-improvement and competitiveness. The recommendation is needed in an E-learning system to filter suitable courses for users when they are facing a massive amount of information in course enrolment. However, due to the complexity of each learning course and the change of user interest, it is challenging to provide accurate recommendations. This paper proposes an E-learning recommender system that combines the recurrent neural network (RNN) and content-based technique to support users in course selection. The content-based techniques are to mine the relationships between courses, and the recurrent neural network is to extract user interests with a series of his/her enrolled courses. The proposed E-learning recommender system framework takes sequential connections into consideration. It intends to provide students with more precise course recommendations. The system is implemented with the Django framework and ElephantSQL cloud database and deployed on the Amazon Elastic Compute Cloud.

Association rules are rules that define relationships between items in sales databases. They have been used primarily to organize relevant products in stores in a way to makes them more visible to consumers, which may increase sales and profits. On the other hand, it has been rarely used in recommender systems where algorithms provide instant recommendations by processing consumers’ interests that are gathered when browsing online. However, the vast amount of information collected from transaction data saved on backup servers is poorly taken advantage of, because it is not connected to the Internet, although interesting and personalized recommendations can be created after finding the collections of most frequent items, or most interesting rules in such databases. In this paper, we do a critique of the existing research on both recommender systems along with showing their drawbacks, and the association rules with detailed explanations on their advantages. Finally, draw up with several solutions for producing high quality as well as accurate recommendations by applying novel combinations of techniques observed in this research area including the association-rules-based recommender systems.

Accurate user preferences and item representations are essential factors for personalized recommender systems. Explicit feedback behaviors, such as ratings and free-text comments, are rich in personalized preference knowledge and emotional evaluation information. It is a direct and effective way to obtain individualized preference and item latent representations from these sources. In this paper, we propose a novel neural model named BERT-RS for personalized recommender systems, which extracts knowledge from textual reviews and user-item interactions. First, we preliminary extract the semantic representation for users and items from the textual comments based on BERT. Next, these semantic embeddings are used for user and item latent representations through three different deep architectures. Finally, we carry out personalized recommendation tasks through the score prediction based on these representations. Compared with other algorithms, BERT-RS demonstrates outstanding experimental performance on the Amazon dataset.

While route recommendation plays a vital role in digital tourism services, conventional methods tend to be inherently complex, as they need to consider several constraints at once: user preferences, geographical and scheduling information, etc. Besides, recent interests from society and institutions have motivated a greater emphasis on sustainability across multiple sectors, including tourism. This work proposes to use reranking techniques as a plain and efficient method to generate personalized routes for users while considering sustainable goals in the procedure. We believe that these results could lead to new research directions regarding sustainable route recommendation techniques.

Cold start and data sparsity issues often plague traditional recommendation systems, as they only rely on rating matrices. Social recommendation systems address these issues by considering users’ social trust relationships. However, trust data is often sparse because users need more information on these supplementary resources. To overcome these challenges, we propose a new recommendation model that models the trust relationship matrix using deep neural networks. In this process, we use the sparse auto-encoder to learn latent features of the user’s trust relationships and use the learned latent features to construct the user’s neighbors. Finally, we combine the original and learned latent social relationships with the matrix factorization model. The experimental results on two datasets have validated the efficiency of this method.

Scalability challenges in recommender systems refer to the difficulties that arise when maintaining systems that can handle growing datasets. On the other hand, state-of-the-art recommender systems are focusing only on increasing the number of transactions (by using evaluation metrics based on rating or ranking). However, the success of a recommender system may be reflected in business metrics, such as increased sales, revenue, user retention, or customer satisfaction. In this chapter, we aim to overcome these two challenges together: “how to define own targets (evaluation metrics) on a recommender system?” and in the meanwhile “how to scale it on big data?”. We proposed a collaborative filtering method called “TOROS: Target Oriented O(n) Recommender System”. TOROS reduces the similarity calculation complexity from O(n²) to O(n) and it has been evaluated on both publicly available datasets and also real-world e-commerce datasets of an e-commerce services provider company Frizbit S.L. We have compared TOROS with state-of-the-art recommender system algorithms and evaluated based on time and space consumption yet. As future work, we also evaluate the efficiency of TOROS in terms of various business-specific targets.

Recommender systems are tools whose objective is to evaluate and filter the great amount of information available to assist the users in their information access processes. In this paper, we present a recommender system for research resources based on fuzzy linguistic modeling, promoting the collaboration between several research groups.

Electronic shops provide an excellent choice to buy without leaving home. Nevertheless, people frequently have problems to find what they look for because of the wide range of items that e-shops offer. Recommender Systems are applications that help people in their searches in e-shops. They deal with information that people provide which is usually related to opinions, tastes and perceptions, and therefore, it is difficult to express them by means of precise numeric scales. Fuzzy linguistic approach provides a better way to express this kind of information. In this contribution we propose a Knowledge Based Recommender System that uses the fuzzy linguistic approach to handle the uncertainty of the human opinions and provides a multigranular context that allows users to utilize the term set that better fits with their degree of knowledge.

Recommender Systems (RSs) are tools for suggesting items that match the preferences and interests for a target user. These systems require the elicitation of customers preferences, which is not always precise because of external factors such as human errors, uncertainty, or vagueness proper of human beings. In RSs, such a problem is known as natural noise (NN) and can bias negatively the recommendations, leading to poor user's experience. The NN management has been addressed in previous works using crisp approaches. This contribution is devoted to a new fuzzy method for managing the uncertainty of NN in a exible and adaptable way for improving recommendations. A case study will show the improvements associated with the proposal.

One of the most promising approaches in the field of Recommender Systems (RSs) is Collaborative Filtering (CF). CF techniques are commonly divided in the two general classes of memory-based and model-based. A wise strategy would be to combine these two methods to increase their performance while leveling out the weakness of each one. Otherwise, the uncertainty pervaded throughout the different steps of the recommendation process should not be ignored. Handling uncertainty is very challenging and important for more reliable and intelligible predictions. That is why, we propose in this paper a new CF approach which combines these two categories under the belief function theory while dealing with the uncertainty pervaded in the prediction process. The effectiveness of our proposal is validated on a real-word data set and compared to state-of-the-art CF approaches under certain and uncertain frameworks.

This talk will introduce some popular recommendation approaches and, in particular, present the recent developments by our Decision Systems and e-Service Intelligence (DeSI) lab in personalised recommender systems and their applications. It consists of three main parts: the models and approaches of recommender systems, the applications of soft computing in recommender systems, and recent developments of recommender system applications in e-business intelligence.