Article
Computer Science, Artificial Intelligence
Aminu Da'u, Naomie Salim, Rabiu Idris
Summary: This paper proposes a RS model that utilizes neural attention techniques to learn adaptive user/item representations and fine-grained user-item interactions, aiming to enhance the accuracy of item recommendation. Experimental results show that the proposed model outperforms existing methods in terms of rating prediction and ranking performances.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Nabila Amir, Fouzia Jabeen, Zafar Ali, Irfan Ullah, Asim Ullah Jan, Pavlos Kefalas
Summary: This survey fills the gap in the literature by summarizing the strengths, weaknesses, and trends of news recommendation models employing DL methods. It also discusses the commonly used datasets, evaluation methods, and implications for researchers in this area.
ARTIFICIAL INTELLIGENCE REVIEW
(2023)
Article
Chemistry, Multidisciplinary
Marcia Barros, Andre Moitinho, Francisco M. Couto
Summary: In this study, a hybrid recommender model is proposed for identifying compounds of interest to scientific researchers, integrating collaborative-filtering algorithms and a new content-based algorithm based on the semantic similarity between chemical compounds. The hybrid model significantly improved the results of collaborative-filtering algorithms by over ten percentage points in most evaluation metrics when evaluated on the implicit dataset CheRM-20 with over 16,000 chemical compounds.
JOURNAL OF CHEMINFORMATICS
(2021)
Article
Computer Science, Artificial Intelligence
Binhui Xie, Shuang Li, Mingjia Li, Chi Harold Liu, Gao Huang, Guoren Wang
Summary: Domain adaptive semantic segmentation aims to achieve satisfactory predictions on an unlabeled target domain by utilizing a supervised model trained on a labeled source domain. We propose SePiCo, a novel one-stage adaptation framework that emphasizes the semantic concepts of individual pixels to improve self-training methods.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Yufeng Wang, Weidong Zhang, Jianhua Ma, Qun Jin
Summary: This study proposes new clustering MAB-based online recommendation methods, ADCB and ADCB+, which address the insufficient feedbacks and dynamics of individual arrival and item popularity in online recommender systems. The experiments consistently show that these two methods outperform existing dynamic clustering-based online recommendation methods in terms of cumulative reward over recommendation rounds and average Click-Through-Rate.
NEURAL PROCESSING LETTERS
(2023)
Review
Computer Science, Interdisciplinary Applications
Fumiya Okubo, Tetsuya Shiino, Tsubasa Minematsu, Yuta Taniguchi, Atsushi Shimada
Summary: In this study, an integrated system is proposed to support learners' reviews. The system uses a review dashboard to recommend adaptive review contents based on individual learners' level of understanding and provide other useful information. Pages of digital learning materials estimated to be insufficiently understood and related webpages are recommended. The experiment showed that the review dashboard was found useful by at least half of the participants for various types of feedback, and it significantly improved learning as indicated by the higher rate of change in quiz scores.
IEEE TRANSACTIONS ON LEARNING TECHNOLOGIES
(2023)
Article
Computer Science, Information Systems
Maryem Rhanoui, Mounia Mikram, Siham Yousfi, Ayoub Kasmi, Naoufel Zoubeidi
Summary: The explosion of information sources has transformed how users access information and give feedback. Modern libraries must adapt to this change by meeting user needs, considering their opinions and preferences, and providing suitable resources.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Pratik K. Biswas, Songlin Liu
Summary: In this paper, a hybrid recommender system that combines collaborative filtering with deep learning is proposed to enhance recommendation performance and overcome the limitations of collaborative filtering. By combining the outputs of collaborative filtering with a deep neural network in a big data processing framework, the proposed system outperforms existing hybrid recommender systems in recommending smartphones to prospective customers.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Mansoureh Ghiasabadi Farahani, Javad Akbari Torkestani, Mohsen Rahmani
Summary: Personalized recommender systems provide preferred services based on user preferences and interests. This study proposes a framework using learning automata to create adaptive user profiling for a personalized recommender system, addressing research gaps and the cold start problem. Experimental results on movie datasets demonstrate that the proposed algorithm outperforms existing approaches in terms of precision, recall, MAE, and RMSE.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Julia Clemente, Hector Yago, Javier De Pedro-Carracedo, Javier Bueno
Summary: This paper proposes a new approach for developing an adaptive competence-based recommender system, using ontological and non-ontological resources to promote improvement in personalized student learning. The importance of flexibility and adaptability in learning modeling and recommendation processes is highlighted.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Mansoureh Ghiasabadi Farahani, Javad Akbari Torkestan, Mohsen Rahmani
Summary: Personalized recommender systems rely on accurate and complete user profiles to provide successful recommendation services. To address the changing interests of users, we propose a learning automata-based algorithm that clusters items and adjusts user interests accordingly. Experimental results demonstrate that our algorithm outperforms other approaches in terms of precision, recall, RMSE, and MAE, and shows acceptable performance for new users.
INFORMATION SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Yuanguo Lin, Shibo Feng, Fan Lin, Wenhua Zeng, Yong Liu, Pengcheng Wu
Summary: This paper introduces a novel course recommendation framework named DARL, which aims to enhance the adaptivity of the recommendation model by automatically capturing users' dynamic interests and adaptively updating the attention weight of courses to improve recommendation accuracy. Empirical experiments on two real-world MOOCs datasets show that DARL significantly outperforms state-of-the-art course recommendation methods in major evaluation metrics.
KNOWLEDGE-BASED SYSTEMS
(2021)
Review
Computer Science, Artificial Intelligence
Tieyuan Liu, Qiong Wu, Liang Chang, Tianlong Gu
Summary: This paper provides a systematic review of deep learning-based recommendation systems in e-learning environments. It introduces the concept and classification of recommendation systems, analyzes existing systems, and presents an overall course recommendation system framework. It focuses on the applications of various deep learning techniques and discusses the flaws in current systems and future research opportunities.
ARTIFICIAL INTELLIGENCE REVIEW
(2022)
Article
Computer Science, Information Systems
Nicolas Torres
Summary: Traditional recommendation systems rely on user ratings as input, while this paper introduces a convolutional neural network architecture that connects user ratings with product images to enhance item recommendations.
Article
Chemistry, Multidisciplinary
Amani Braham, Maha Khemaja, Felix Buendia, Faiez Gargouri
Summary: Design patterns are acknowledged as a solution to recurring design problems. This study contributes by developing a recommender system for selecting relevant design patterns and validating its potential impact on acceptance intention.
APPLIED SCIENCES-BASEL
(2021)
Article
Engineering, Electrical & Electronic
Valentina Gatteschi, Alberto Cannavo, Fabrizio Lamberti, Lia Morra, Paolo Montuschi
Summary: Aggressive driving is a major cause of fatal crashes, but correctly identifying aggressive driving events is still challenging. This study compares the performance of various algorithms for aggressive driving event detection and explores the potential of smartphones as a replacement for black-box devices.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2022)
Article
Computer Science, Hardware & Architecture
Wahab Almuhtadi, Fabrizio Lamberti
Summary: This article summarizes the efforts made by CTSoc in establishing the Technical Activities area and its technical committees over the past two years.
IEEE CONSUMER ELECTRONICS MAGAZINE
(2022)
Article
Computer Science, Interdisciplinary Applications
Davide Calandra, Federico De Lorenzis, Alberto Cannavo, Fabrizio Lamberti
Summary: This study investigates the use of a Virtual Reality Training Simulation (VRTS) as a complement to traditional video-based training methods in the context of forest firefighting. Results show that the use of VRTS improves trainees' procedural learning and perceived quality of the overall learning experience.
Article
Green & Sustainable Science & Technology
Gema Millan, Yassine Rqiq, Erudino Llano, Victor Ballestin, Lisa Neusel, Antoine Durand, Josephine Troeger, Fabrizio Lamberti, Federico De Lorenzis, Maurizio Repetto
Summary: Energy efficiency requirements in Europe are determined by the Energy Efficiency Directive, with energy audits as a systematic procedure. However, the current regulation only applies to non-SMEs, posing challenges for small businesses. Therefore, raising awareness and providing training is crucial in promoting energy performance improvement among small enterprises.
Review
Biology
Fabio Garcea, Alessio Serra, Fabrizio Lamberti, Lia Morra
Summary: Recent advances in Deep Learning have benefited greatly from larger and more diverse training sets, but collecting large datasets for medical imaging remains a challenge. Data augmentation techniques allow for expansion of available training data without collecting new samples. Different data augmentation strategies perform differently depending on input nature and visual tasks. This literature review investigates the use and impact of data augmentation strategies in the medical domain and suggests potential avenues for future research based on comprehensive analysis of over 300 articles published in recent years (2018-2022).
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
Article
Computer Science, Interdisciplinary Applications
Federico De Lorenzis, Filippo Gabriele Prattico, Maurizio Repetto, Enrico Pons, Fabrizio Lamberti
Summary: This study explores the application of Virtual Reality Training Systems (VRTS) in the field of energy management. The results showed that VRTS improves participants' performance and that learning by teaching is more effective in practical learning.
COMPUTERS IN INDUSTRY
(2023)
Article
Multidisciplinary Sciences
Fabio Garcea, Giacomo Blanco, Alberto Croci, Fabrizio Lamberti, Riccardo Mamone, Ruben Ricupero, Lia Morra, Paola Allamano
Summary: This research proposes a deep learning-based solution for automatically detecting wet road events using road-side surveillance cameras. The technique employs a convLSTM model to detect subtle changes in road appearance and a contrastive self-supervised framework tailored to surveillance camera networks. Experimental results show the effectiveness of the proposed techniques in improving wet road event detection and reducing false positive alarms.
SCIENTIFIC REPORTS
(2022)
Article
Computer Science, Artificial Intelligence
Alberto Cannavo, Antonio Castiello, F. Gabriele Prattico, Tatiana Mazali, Fabrizio Lamberti
Summary: This paper investigates the impact of different perspectives (first-person perspective and external perspective) on narrative engagement in cinematic virtual reality (CVR) films and finds that the first-person perspective can enhance overall narrative engagement and presence.
Article
Computer Science, Cybernetics
Alberto Cannavo, Emanuele Stellini, Congyi Zhang, Fabrizio Lamberti
Summary: Creating facial animations using 3D computer graphics is a laborious task. The use of blendshapes is common, but has drawbacks such as the need to memorize mappings and limitations in expressiveness. This article proposes a virtual reality-based interface that uses sketches for direct manipulation of blendshapes, addressing these issues.
INTERNATIONAL JOURNAL OF HUMAN-COMPUTER INTERACTION
(2023)
Article
Computer Science, Information Systems
Alberto Butera, Valentina Gatteschi, Filippo Gabriele Prattico, Daniela Novaro, Deborah Vianello
Summary: Recently, the automotive industry has witnessed disruptive innovations like self-driving cars and hybrid/electric engines. However, certain operations such as second-hand vehicle trades still follow traditional methods, leading to trust issues between buyers and sellers. Studies indicate that odometer fraud alone costs around 8.9 billion euros annually. To address this, blockchain technology is proposed to transparently store a vehicle's history and create a decentralized second-hand vehicle market using Non-Fungible Tokens (NFTs) for automatic ownership transfers. An architecture and practical implementation of a Decentralized Application (Dapp) are presented alongside discussions on security, costs, and future developments.
Proceedings Paper
Computer Science, Artificial Intelligence
Simone Martone, Francesco Manigrasso, Fabrizio Lamberti, Lia Morra
Summary: Semantic image interpretation can greatly benefit from combining sub-symbolic distributed representation learning with higher-level abstraction reasoning. Logic Tensor Networks (LTNs) replace traditional training sets with knowledge bases of fuzzy logical axioms, and they can learn to satisfy the knowledge base through differentiable operators. PROTOtypical Logic Tensor Networks (PROTO-LTN) extend LTNs by grounding abstract concepts as parameterized class prototypes, reducing the required parameters for grounding the knowledge base. The proposed implementation shows promising results in zero-shot learning scenarios.
2022 26TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR)
(2022)
Article
Computer Science, Information Systems
Luca Piano, Fabio Garcea, Valentina Gatteschi, Fabrizio Lamberti, Lia Morra
Summary: In this article, we review existing methods for dataset drift detection, discuss their applicability to deep neural networks, and experiment on a practical case study related to semistructured document analysis.
Proceedings Paper
Computer Science, Interdisciplinary Applications
Alberto Cannavo, Valentina Gatteschi, Luca Macis, Fabrizio Lamberti
Summary: This paper introduces a system that can automatically generate computer graphics videos by processing text descriptions and associated images. The system combines natural language processing and image analysis to extract information and visually represent it with 3D animations. The system has achieved promising results in a specific use case focused on printer maintenance.
EXTENDED REALITY, XR SALENTO 2022, PT I
(2022)
Proceedings Paper
Engineering, Electrical & Electronic
Davide Calandra, Filippo Gabriele Prattico, Fabrizio Lamberti
Summary: Speech-based navigation techniques in VR training can enhance user experience and avoid the limitations of handheld controllers. This study compares different speech-based navigation techniques in a virtual environment to evaluate their usability and performance.
2022 IEEE 21ST MEDITERRANEAN ELECTROTECHNICAL CONFERENCE (IEEE MELECON 2022)
(2022)