The author proposes the Meta-Split Network (MSN) to address the challenges of limited-stock products in C2C e-commerce platforms by segmenting user history based on stock levels and applying a unique meta-learning approach for low-stock items.
MSNet improves CTR prediction for limited-stock products in C2C e-commerce.
Die Meta-Split-Netzwerk (MSN) Methode verbessert die CTR-Vorhersagen für begrenzte Lagerprodukte in C2C-E-Commerce-Plattformen.