Online inference and learning for nonsymmetric determinantal point processes

US12288237B2 · kind B2 · utility

Assignee

• Adobe Inc. · US

Inventors

• Ryan A. Rossi · Mountain View, US
• Aravind Reddy Talla · Evanston, US
• Zhao Song · Princeton, US
• Anup Rao · San Jose, US
• Tung Mai · San Jose, US
• Nedim Lipka · Campbell, US
• Gang Wu · Zhejiang, CN
• Eunyee Koh · San Jose, US

Key dates

• Filing date: May 12, 2022
• Grant date: Apr 29, 2025
• Expiry date: Dec 21, 2042

Classification

• Technology area (CPC G): Physics
• CPC primary: G06Q30/0643
• WIPO field: IT methods for management
• WIPO sector: Electrical engineering

Abstract

Embodiments provide systems, methods, and computer storage media for a Nonsymmetric Determinantal Point Process (NDPPs) for compatible set recommendations in a setting where data representing entities (e.g., items) arrives in a stream. A stream representing compatible sets of entities is received and used to update a latent representation of the entities and a compatibility distribution indicating likelihood of compatibility of subsets of the entities. The probability distribution is accessed in a single sequential pass to predict a compatible complete set of entities that completes an incomplete set of entities. The predicted complete compatible set is provided a recommendation for entities that complete the incomplete set of entities.