prior-art-search

A prior art search identifies existing patents and academic literature relevant to an invention to evaluate its novelty and inventive-step risk. It extracts technical features and matches them against existing publications to flag potential patentability issues.

To find prior art, provide an invention description to extract core concepts, problems, and technical features. The system searches Google Patents, Espacenet, Google Scholar, and arXiv using English and Chinese queries, then outputs a structured report with overlap risk labeling.

Yes, prior art search supports freedom-to-operate pre-checks by analyzing extracted invention concepts against existing patents. It generates an actionable report that includes a preliminary freedom-to-operate assessment and labels overlap risk.

Yes, the prior art search queries patent and literature databases using both English and Chinese queries when needed. It searches platforms like Google Patents, Espacenet, Google Scholar, and arXiv to ensure comprehensive coverage of relevant prior publications.

The best way to classify invention features is to break the invention into its core concept, problem, and technical features, then predict IPC/CPC classes. This concept-driven extraction structures the search queries and analyzes overlaps by relevance and anticipation risk.

No, prior art search reports are strictly non-fabricated. The system only outputs actual patents and academic publications found during the database search, providing a prioritized, structured report with genuine relevance and overlap risk labeling.