Deep Networks as a Semantic Platform for Modeling User Behavior Data

• Forschungsstelle: INNOSUISSE
• Projektnummer: 18785.1 PFES-ES
• Projekttitel Englisch: Deep Networks as a Semantic Platform for Modeling User Behavior Data

Kurzbeschreibung (Deutsch)

Deep Networks as a Semantic Platform for Modeling User Behavior Data

Kurzbeschreibung (Englisch)

Deep Networks as a Semantic Platform for Modeling User Behavior Data

Abstract (Deutsch)

We propose to build a software platform for modeling, integrating, and utilizing data collected from user interactions in an online setting. Deep networks are used as a modeling framework along with existing highperformance computational frameworks. Data coming from different modalities are semantically integrated into a common embedding space. This supports predictive tasks from personalization and content recommendation to ads targeting and optimized bidding.

Abstract (Englisch)

We propose to build a software platform for modeling, integrating, and utilizing data collected from user interactions in an online setting. Deep networks are used as a modeling framework along with existing highperformance computational frameworks. Data coming from different modalities are semantically integrated into a common embedding space. This supports predictive tasks from personalization and content recommendation to ads targeting and optimized bidding.