Identification of domain-relevant patents via weakly supervised deep learning

Key Points

• The analysis reveals significant improvements in patent identification accuracy with weakly supervised deep learning techniques.
• A 30% increase in identification performance was observed when compared to traditional methods in domain analysis.
• Observational analysis uses deep learning models to assess patent relevance based on labeled and unlabeled data.
• This approach may enable more efficient technology assessments, offering a new avenue for innovation tracking.