Personalized e-commerce product recommendations
This whitepaper provides a guide on implementing a recommender system in e-commerce, covering the selection of the right system, ML models, data pipelines, and training algorithms.
Fighting with biases: How to measure diversity in streaming media
This article discusses the importance of diversity in recommendation systems and proposes a method for automatically estimating video diversity based on gender, ethnicity, and age. The method involves face detection, gender and ethnicity classification, and batch processing for efficient video analysis.
Understanding search query intent with deep learning
This article discusses the challenge of underspecified queries in e-commerce search and how deep learning NLP models can be used to understand customer intent and provide relevant results, leading to improved revenue per session.
How deep learning improves recommendations for 80% of your catalog
This article discusses how deep learning can be used to improve behavior-driven product recommendations, even for products with limited customer interaction data, by training a neural network to predict latent features based on product attributes and content.
