声明
Abstract
摘要
List of Contents
Chapter 1 Introduction
1.1 Recommender Systems:Background
1.2 Motivation
1.3 Problem Statement
1.4 Research Objective
1.5 Contributions
1.6 Thesis Outline
Chapter 2 Basics and Related Work
2.1 Formalization of Recommendation Problem
2.2 Users’ and Items’ Profiles
2.3 Classification of Recommender Systems
2.3.1 Collaborative Filtering (CF) Recommender Systems
2.3.2 Content-Based Filtering (CBF) recommender systems
2.3.3 Knowledge-Based (Kb) Recommender Systems
2.3.4 Demographic-Based (DM) Recommender Systems
2.3.5 Hybrid Recommender Systems
2.3.6 Other Types of Recommender Systems
2.4 Related Work
2.5 Our Proposed Work
Chapter 3 Hybrid Recommender Systems
3.1 Introduction
3.2 Background
3.3 Naive Bayes Classifier
3.4 Support Vector Machines (SVM)
3.5 Combining the Item-Based CF and Classification Approaches for Improved Recommendations
3.6 Combining the Item-Based CF and the Naive Bayes Classifier (SwitchRecNBCF)
3.8 Combining the Item-Based CF and the SVM Classifier (SwitchRecsvmcf)
Chapter 4 Evaluation and Results
4.1 Datasets
4.2 Getting Additional Features about Movies
4.3 Evaluation Metrics
4.3.1 Mean Absolute Error (MAE) and Related Metrics
4.3.2 Receiver Operating Characteristic (ROC)-Sensitivity
4.3.3 Precision,Recall,and F1 Measure
4.3.4 Coverage
4.3.5 Other Metries
4.3.6 Evaluation From The User’s Point of View
4.4 Presenting Recommendations to Users
4.5 Evaluation Methodology
4.6 Results and Discussion
4.6.1 Learning the Optimal System Parameters
4.6.2 Performance Evaluation with Other Algorithms
4.6.3 Eliminating Over-Specialization Problem
Chapter 5 Conclusions and Future Works
5.1 Conclusions
5.2 Future Works
References
Acknowledgments