Recommender Systems in Fashion and Retail (Lecture Notes in Electrical Engineering, 734)

Contents
Fashion Understanding
The Importance of Brand Affinity  in Luxury Fashion Recommendations
	1 Introduction
	2 Related Works
	3 Methodology
		3.1 Data Collection
		3.2 Data Preparation
		3.3 Brand Affinity Modelling
		3.4 Boosting Recommendations with Brand Affinity Information
	4 Offline Experiment
		4.1 Offline Setup and Model Selection
		4.2 Results and Discussion of Offline Evaluation
	5 Online Experiment
		5.1 Online Setup
		5.2 Results and Discussion of Online Evaluation
	6 Conclusion and Future Work
	References
Probabilistic Color Modelling  of Clothing Items
	1 Introduction
	2 Color Modelling and Extraction
		2.1 Mathematical Modelling
		2.2 Clothing Instance Segmentation
		2.3 Extracting the Main Colors
		2.4 Merging Pure Colors that Have Different Tints/shades
		2.5 Probabilistic Modelling
		2.6 Color Names
	3 Results
		3.1 The Effect of Number of Clusters
		3.2 Comparison with Color Extraction Tools
		3.3 Color Distributions of Fashion Data
		3.4 Fashion Color Trend
	4 Discussion and Conclusion
		4.1 Gaussian Mixture Model Versus K-Means
		4.2 Probabilistic Color Model
		4.3 Making Use of Color Extraction in Fashion
		4.4 Color Perception and Evaluation
		4.5 Future Prospects
	References
User Aesthetics Identification for Fashion Recommendations
	1 Introduction
	2 Related Work
	3 Methodology
		3.1 General Users Statistics Model
		3.2 Image Embedding Model
		3.3 Word Embedding Model
	4 Experiments and Results
		4.1 Dataset and Evaluation
		4.2 Discussion
	5 Conclusion and Future Work
	References
Sizing and Fit in Online Fashion
Towards User-in-the-Loop Online Fashion Size Recommendation with Low Cognitive Load
	1 Introduction
	2 Complexity of the Size and Fit Problem at Scale
	3 Prior Work
	4 Size Recommendation Without Order History
	5 Experimental Results and Discussion
		5.1 Hot-Start and Cold-Start Performances
		5.2 Impact of Brand Size Offsets
		5.3 Customer Coverage
		5.4 Hybrid Recommendation Systems
		5.5 Minimizing Customers\' Cognitive Load
		5.6 Performance in Production
		5.7 Leveraging Customer Data for Hot-Start Recommendation
		5.8 Summary
	6 Conclusion
	References
Attention Gets You the Right Size and Fit in Fashion
	1 Introduction
	2 Related Work
	3 Proposed Approach
		3.1 Inputs and Embeddings
		3.2 Encoder and Decoder Layers
	4 Experimental Setup
		4.1 Large-Scale Anonymized Data
		4.2 Training, Validation and Test Samples
		4.3 Experimental Details
	5 Results and Discussion
		5.1 Overall Performance Comparison
		5.2 Cross-Category Performance
		5.3 Attention Adapts to Changes in the History
		5.4 Performance on Multi-user Accounts
		5.5 Online Performance
	6 Results on Public Datasets
	7 Conclusion
	References
Combining Fashion
The Ensemble-Building Challenge for Fashion Recommendation: Investigation of In-Home Practices and Assessment of Garment Combinations
	1 Introduction
	2 In-Home Outfit Building Strategies
		2.1 Methods
		2.2 Results and Discussion
	3 Assessing Garment Combinations
		3.1 Method
		3.2 Results and Discussion
	4 Conclusions
	References
Outfit Generation and Recommendation—An Experimental Study
	1 Introduction
	2 Related Work
	3 Algorithms
		3.1 Item Compatibility
		3.2 Personalized Outfit Generation Algorithms
	4 Experiments
		4.1 Datasets
		4.2 Item Representation
		4.3 Non-Personalized Models
		4.4 Personalized Outfit Generation
	5 Conclusions and Future Work
	References
Understanding Professional Fashion Stylists’ Outfit Recommendation Process: A Qualitative Study
	1 Introduction
	2 Research Methodology
		2.1 Interview Questions
		2.2 Participants
		2.3 Data Collection
		2.4 Data Analysis
	3 Findings
		3.1 What Factors Influence How Professional Stylists Make Decisions on Outfit Suitability and Choice?
		3.2 Do the Outfit Suitability and Choice Factors Change if They Are Integrating the Existing Wardrobe or not?
	4 Discussion
		4.1 Client Features
		4.2 Garment Features
		4.3 Implications for Recommender Systems Design
	5 Conclusions
	References