Data-fusion-based Patent Valuation: Discovering the Market Need

As innovation-driven development becomes a global consensus, patent valuation has received increasing attention.However, the study of automatic patent valuation is still in its infancy, and traditional models cannot effectively estimate patent value.Since patent value is inherently linked to market needs, understanding market needs is crucial for patent valuation.Current challenges in patent valuation include: 1) Existing patent valuation models neglect market needs; 2) Few disclosed patent value data are available.To solve these issues, we propose a datafusion-based patent need valuation model to discover the market needs of patents, and calculate the need values of patents.Our model includes 2 modules: 1) Discovering market needs from annual reports and patents of listed companies and constructing a value heterogeneous network; 2) Calculating the patent need values.In the experiments, we use our model to automatically extract the market needs, and calculate the need values.