Article
Computer Science, Artificial Intelligence
F. Ortega, J. Mayor, D. Lopez-Fernandez, R. Lara-Cabrera
Summary: CF4J 2.0 is a framework designed for research experiments based on collaborative filtering, with features like implemented algorithms, quality measures, parallel execution, and abstract classes for developers to customize. The new version focuses on simple deployment, reproducible science, hyper-parameter optimization, data analysis, and community openness as an open-source project.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Information Systems
Fahrettin Horasan, Ahmet Hasim Yurttakal, Selcuk Gunduz
Summary: Collaborative filtering is a technique that considers the common characteristics of users and items in recommendation systems. Matrix decompositions, such as SVD and NMF, are widely used in collaborative filtering. In this study, a technique called T-ULVD was used to improve the accuracy and quality of recommendations. Experimental results showed that T-ULVD achieved better results compared to NMF and performed as well as or even better than SVD. This study may provide guidance for future research on solving the cold-start problem and reducing sparsity in collaborative filtering based recommender systems.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2023)
Article
Computer Science, Information Systems
Lianghao Xia, Chao Huang, Yong Xu, Huance Xu, Xiang Li, Weiguo Zhang
Summary: As deep learning techniques are applied to real-world recommendation tasks, many deep neural network based Collaborative Filtering (CF) models have been developed. However, most existing models are not well equipped to handle missing data. In this paper, we propose a Collaborative Reflection-Augmented Autoencoder Network (CRANet) that can leverage both observed and unobserved user-item interactions to improve recommendation performance. We also introduce a robust joint training algorithm using regularization-based tied-weights. Experimental results show that debiasing negative signals improves the performance compared to state-of-the-art techniques.
ACM TRANSACTIONS ON INFORMATION SYSTEMS
(2022)
Article
Computer Science, Information Systems
Le Nguyen Hoai Nam
Summary: This paper focuses on the rating prediction phase in memory-based collaborative filtering and improves the prediction accuracy by optimizing an objective function. Experimental results demonstrate that the proposed method outperforms others, especially when the number of selected neighbors is small to medium.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Hao Tang, Guoshuai Zhao, Xuxiao Bu, Xueming Qian
Summary: The recommendation system is an important technology in the era of Big Data. Current methods have integrated side information to alleviate the sparsity problem, but not all side information can be obtained with high quality. By proposing the DMGCF model and dynamically evolving multi-graph collaborative filtering, the approach successfully mines and reuses side information, as demonstrated by experimental results.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Rabeh Ravanifard, Abdolreza Mirzaei, Wray Buntine, Mehran Safayani
Summary: Listwise collaborative filtering algorithms are gaining interest for their efficiency and accuracy in recommendation systems. A Bayesian graphical model called CALCF is proposed in this work to integrate text information into listwise CF, achieving better performance in recommendation tasks.
Article
Computer Science, Artificial Intelligence
Jesus Bobadilla, Fernando Ortega, Abraham Gutierrez, Angel Gonzalez-Prieto
Summary: The research introduces a method to incorporate stochasticity into deep learning models using variational autoencoders, aiming to improve the performance of recommender systems. By introducing variational techniques in the latent space, this approach can be applied as a plugin to current and future models, demonstrating superior performance in experiments.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Jesus Bobadilla, Abraham Gutierrez, Raciel Yera, Luis Martinez
Summary: This paper proposes a method based on Generative Adversarial Networks (GANs) to generate collaborative filtering datasets with specific features. The method uses dense, short, and continuous embeddings for faster and more accurate learning compared to traditional methods.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Pham Minh Thu Do, Thi Thanh Sang Nguyen
Summary: This paper proposes a novel semantic-enhanced Neural Collaborative Filtering (NCF) model for movie rating prediction and recommendation tasks. By building a semantic knowledge base and user behavior analytic model, combined with user preferences and recommendation model, the proposed model shows better recommendation performance in experiments.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Information Systems
Alvaro Gonzalez, Fernando Ortega, Diego Perez-Lopez, Santiago Alonso
Summary: Recommender Systems, an essential tool in streaming and marketplace systems, have been found to exhibit clear bias and unfairness towards minorities and underrepresented groups. This paper analyzes the demographic characteristics of a gold standard dataset and proposes Soft Matrix Factorization (SoftMF) to balance predictions and reduce existing inequality.
Article
Computer Science, Information Systems
Sofia Bourhim, Lamia Benhiba, M. A. Janati Idrissi
Summary: This article presents a graph-based recommender system that improves accuracy by considering the behavior of communities and the interaction between sub-groups in the network. Experimental results show significant improvements compared to other deep learning models, and the system has the potential to extract deep relationships for better recommendations.
Article
Chemistry, Multidisciplinary
Abebe Tegene, Qiao Liu, Yanglei Gan, Tingting Dai, Habte Leka, Melak Ayenew
Summary: A collaborative recommender system based on a latent factor model has achieved significant success in the field of personalized recommender systems. However, the latent factor model suffers from sparsity problems and limited ability in extracting non-linear data features, resulting in poor recommendation performance. In this paper, we propose a dual deep learning and embedding-based latent factor model that considers dense user and item feature vectors to overcome these problems and improve rating prediction performance.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Information Systems
Keren Gaiger, Oren Barkan, Shir Tsipory-Samuel, Noam Koenigstein
Summary: Collaborative filtering methods often represent users as static latent vectors, but user behavior and interests can change dynamically in response to recommended items. In order to address this issue, the Attentive Item2Vec++ (AI2V++) model is introduced, which adaptively adjusts the user representation based on the recommended item. AI2V++ utilizes a novel context-target attention mechanism to capture different characteristics of the user's historical behavior towards potential recommendations. Furthermore, the model's interpretability and explainability are improved through the analysis of neural-attentive scores. The proposed approach outperforms state-of-the-art baselines across multiple accuracy metrics on five publicly available datasets.
Review
Computer Science, Information Systems
Fethi Fkih
Summary: This paper provides an in-depth review of similarity measures used in collaborative filtering-based recommender systems. Through experimental studies, the performance of different measures is compared, and important conclusions are drawn. Evaluation results show that different similarity measures have different suitability in user-based and item-based recommendations.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Information Systems
Bushra Alhijawi, Ghazi Al-Naymat, Nadim Obeid, Arafat Awajan
Summary: This study introduces a novel prediction mechanism, inheritance-based prediction (INH-BP), for collaborative filtering recommender systems, which customizes the predictor based on user context. Experimental results demonstrate the superiority of INH-BP in accuracy and its ability to address cold start and sparsity issues.
INFORMATION SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Chhavi Rana, Sanjay Kumar Jain
SWARM AND EVOLUTIONARY COMPUTATION
(2014)
Article
Computer Science, Information Systems
Manvi Breja, Sanjay Kumar Jain
Summary: Why-type non-factoid questions are more complex and difficult to answer compared to factoid questions, with understanding user intent and context being crucial. This paper proposes a classification model trained on lexical, syntactic, and semantic features to classify why-type questions, achieving 81% accuracy in determining question type and 76.8% accuracy in determining answer type with support vector machine outperforming baseline by 14.6%.
INTERNATIONAL JOURNAL OF INFORMATION TECHNOLOGY AND WEB ENGINEERING
(2021)
Article
Computer Science, Information Systems
Manvi Breja, Sanjay Kumar Jain
Summary: Understanding the actual need of users in answering non-factoid why-questions is crucial. This paper analyzes different types of why-questions and proposes algorithms to determine the focus and reformulate the questions for retrieving expected answers. The results show that the precision is 89% for informational why-questions and 48% for opinionated why-questions.
INTERNATIONAL JOURNAL OF INFORMATION RETRIEVAL RESEARCH
(2022)
Article
Computer Science, Information Systems
Manvi Breja, Sanjay Kumar Jain
Summary: This paper explores the extraction, re-ranking, and validation of appropriate answers for non-factoid why-type questions. By utilizing deep learning frameworks and investigating lexico-syntactic, semantic, and contextual query-dependent features, the authors develop a model that uses Ensemble ExtraTreesClassifier to weigh the importance of these features. The model achieves a mean reciprocal rank of 0.64 by finding the highest-ranked answer with the most feature similarity to the question and validating the answer by matching the answer type.
JOURNAL OF CASES ON INFORMATION TECHNOLOGY
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Amit Mishra, Sanjay Kumar Jain
COMPUTATIONAL LINGUISTICS AND INTELLIGENT TEXT PROCESSING (CICLING 2015), PT II
(2015)
Article
Engineering, Multidisciplinary
Mohd Sadiq, S. K. Jain
INTERNATIONAL JOURNAL OF SYSTEM ASSURANCE ENGINEERING AND MANAGEMENT
(2015)
Article
Computer Science, Information Systems
Mrityunjay Singh, S. K. Jain
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2015)
Article
Engineering, Multidisciplinary
Mohd Sadiq, S. K. Jain
INTERNATIONAL JOURNAL OF SYSTEM ASSURANCE ENGINEERING AND MANAGEMENT
(2014)
Article
Computer Science, Information Systems
Chhavi Rana, Sanjay Kumar Jain
SOCIAL NETWORK ANALYSIS AND MINING
(2014)
Proceedings Paper
Computer Science, Artificial Intelligence
Saurabh Parmar, S. K. Jain, Gagandeep Kaur, Anand Kumar
GLOBAL TRENDS IN COMPUTING AND COMMUNICATION SYSTEMS, PT 1
(2012)
Proceedings Paper
Computer Science, Hardware & Architecture
Chhavi Rana, Sanjay Kumar Jain
DISTRIBUTED COMPUTING AND INTERNET TECHNOLOGY
(2012)
