Table of contents

• Abstract
• Dedications
• Acknowledgements
• Abbreviations
• Introduction
• Contributions
• Outline
• Part 1 - Recommender Systems
• Chapter 1 Historical development and related work
• Information Retrieval
• Information Filtering
• Collaborative Filtering
• Recommendation techniques
• User based Collaborative Filtering
• Item based Collaborative Filtering
• Similarity measures
• Pearson's correlation coefficient
• Cosine angle
• Adjusted Cosine
• Mean squared difference
• Recommendations
• Prediction algorithms
• Ranking algorithms
• Content Based Filtering
• Vector Space Model
• Hybrid Filters
• Recommender Systems
• Recommender systems in e-commerce
• Trust in recommender systems
• Recommendation approaches
• Manual approach
• Statistical approach
• Content based filtering approach
• Collaborative filtering approach
• Hybrid approach
• Problems and challenges of the recommendation approaches
• The collaborative filtering approach - weaknesses and strengths
• The content based filtering approach - weaknesses and strengths
• Case studies of recommender systems in the movie domain
• MovieLens
• Recommendations
• Predictions
• How it works
• Does it work?
• Browsing
• Searching
• Other features
• Internet Movie Database
• Browsing
• Searching
• Recommendations
• How it works
• Does it work?
• Other features
• All Movie Guide
• Browsing
• Searching
• Recommendations
• How it works
• Does it work?
• Amazon
• Browsing
• Searching
• Recommendations
• How it works
• Does it work?
• Other features
• Recommendations for the movie The Thing
• Conclusions
• Patents
• Where are we today?
• Chapter 2 Implementation and evaluation of common recommendation techniques
• Recommendation profiles
• Information analysis
• Profile representation
• User rating models
• User transaction models
• Item rating models
• Item transaction models
• Item attribute models
• Recommendation techniques
• Baseline Statistical Filtering (BASELINESF)
• User mean prediction algorithm
• Item mean prediction algorithm
• Random prediction algorithm
• Population deviation from mean prediction algorithm
• User based Collaborative Filtering (UCF)
• Pearson similarity algorithm
• Weighted deviation from mean prediction algorithm
• Random Neighbors User based Collaborative Filtering (RNUCF)
• Trust User based Collaborative Filtering (TRUSTUCF)
• Common Items Prioritizing User based Collaborative Filtering (CIPUCF)
• Item based Collaborative Filtering (ICF)
• Adjusted cosine similarity algorithm
• Average rating prediction algorithm
• Weighted sum prediction algorithm
• Random Neighbors Item based Collaborative Filtering (RNICF)
• Personalized Content Based Filtering (PCBF)
• Common attribute similarity algorithm
• Cosine similarity algorithm
• Same decade similarity algorithm
• Top-N User based Collaborative Filtering (TOPNUCF)
• Top-N Item based Collaborative Filtering (TOPNICF)
• Top-N Personalized Content Based Filtering (TOPNPCBF)
• Top-N Content Based Filtering (TOPNCBF)
• Similar items ranking list algorithm
• Transaction based Top-N User based Collaborative Filtering (TBTOPNUCF)
• Cosine similarity algorithm (Simplified transaction case)
• Frequent items ranking list algorithm
• Transaction based Top-N Item based Collaborative Filtering (TBTOPNICF)
• Evaluation of recommender systems
• Evaluation data
• Evaluation protocols
• Splitting evaluation data
• Prediction evaluation metrics
• Coverage
• Average User Coverage
• Mean Absolute Error (MAE)
• Average User MAE
• Normalized Mean Absolute Error (NMAE)
• Average User NMAE
• Mean Squared Error (MSE)
• Average User MSE
• Root Mean Squared Error (RMSE)
• Average User RMSE
• Correctness
• Ranking evaluation metrics
• Coverage
• Recall
• Average User Recall
• Precision
• Average User Precision
• (Breeze's / Halflife) Utility
• Average User (Breeze's / Halflife ) Utility
• F1
• Average User F1
• Average First Hit Position
• Evaluation data
• Consistent MovieLens rating dataset (CML)
• Discshop rating dataset (DS)
• Discshop rating dataset 2 (DS2)
• Discshop transaction dataset (DS-T)
• Discshop attribute dataset (DS-A)
• IMDb attribute dataset (IMDB)
• Evaluation results
• Accuracy of predictions
• BASELINESF Technique
• UCF Technique
• RNUCF Technique
• TRUSTUCF Technique
• CIPUCF Technique
• ICF Technique
• RNICF Technique
• PCBF Technique
• Relevance of Top-N ranking lists
• TOPNUCF Technique
• TOPNICF Technique
• TOPNPCBF Technique
• TOPNCBF Technique
• TBTOPNUCF Technique
• TBTOPNICF Technique
• Summary and conclusions
• Part 2 - Self-Organization applied to Recommender Systems
• Chapter 3 Artificial neural networks
• The brain
• The biological neuron
• The artificial neuron
• Formal definition of artificial neural networks
• Artifical neural network structures and learning algorithms
• Feedforward networks
• Recurrent networks
• Learning Rules
• Error-correction learning
• Hebbian Learning
• Competitive learning
• Learning paradigms
• Learning protocols
• A brief history of artificial neural networks
• Chapter 4 Self-Organizing Maps
• Brain maps
• Requirements for self-organization
• The Mexican Hat
• The SOM paradigm
• The incremental SOM algorithm
• Selection of the best matching node ("Winner-takes-all")
• Adaptation (Updating of the weight vectors)
• Selection of parameters
• Incomplete data
• Quality measure of the SOM
• Average quantization error
• Topographic error
• Summary of the SOM algorithm
• The Dot-Product SOM algorithm
• The batch SOM algorithm
• The SOM as a clustering and projection algorithm
• Vector quantization
• Vector projection
• Linear projection
• Non-linear projection
• Comparison of the SOM to other VQ and VP algorithms
• Visualizing the SOM
• Properties of the SOM
• Approximation of the input space
• Topological ordering
• Illustration of the ordering principle implemented by the SOM
• Density matching
• Theoretical aspects of the SOM algorithm
• SOM Toolbox for Matlab
• SOM based applications
• Applications of the SOM in recommender systems
• Chapter 5 Implementation and evaluation of SOM based recommendation techniques
• SOM implementation
• SOM based recommendation techniques
• SOM User based Collaborative Filtering (SOMUCF)
• SOM Item based Collaborative Filtering (SOMICF)
• Item SOM User based Collaborative Filtering (ISOMUCF)
• Model Predicting SOM User based Collaborative Filtering (MPSOMUCF)
• SOM Goodness Collaborative Filtering (SOMGOODNESSCF)
• Top-N SOM User based Collaborative Filtering (TOPNSOMUCF)
• Top-N SOM Item based Collaborative Filtering (TOPNSOMICF)
• Top-N SOM Goodness Collaborative Filtering (TOPNSOMGOODNESSCF)
• Evaluation data
• Evaluation results
• Accuracy of predictions
• SOMUCF Technique
• SOMICF Technique
• ISOMUCF Technique
• MPSOMUCF Technique
• SOMGOODNESSCF Technique
• Relevance of Top-N ranking lists
• TOPNSOMUCF Technique
• TOPNSOMICF Technique
• TOPNSOMGOODNESSCF Technique
• Conclusions
• Part 3 - The recommendation interface problem
• Chapter 6 Trust in recommender systems
• Transparency of recommendations
• Explanation of recommendations
• Interactive design of recommender systems
• Chapter 7 Visual recommendations
• Chapter 8 MOVSOM - State of the art
• MOVSOM map
• MOVSOM interface
• MOVSOM architecture
• Challenges
• Introduction of new movies and new users
• Retraining the SOM
• Chapter 9 Conclusions
• Chapter 10 Future work
• Bibliography
• Index

Copyright © 2005-2025 MOVSOM™ Five By Five