高航专利评价分析数据

根据待分析专利的法律状态、著录项目特性、说明书全文特性等信息，从专利估值模型的函数数据库中获取对应的估值函数，计算获得待分析专利的估算价值。估值模型是基于大数据训练获得的专利价值计算模型，由函数数据库和参数数据库组成。训练过程主要是基于专利的已有特征数据，进行无效数据剔除、去重，根据特征分组分别进行缺失数据修补、数据统一，然后对特征分组进行数据标注，主要的特征分组包括专利需求匹配度、专利分类匹配度、市场意愿度、授权时效、专利文本质量等。由此训练模型得到的专利价值估值，可用于企业专利布局分析、专利价值分析。

Based on information such as the legal status, bibliographic characteristics, and full specification characteristics of the patent to be analyzed, the corresponding valuation function is retrieved from the function database of the patent valuation model to calculate the estimated value of the patent. The patent valuation model is a patent value calculation model trained on big data, which comprises a function database and a parameter database. The training workflow primarily includes: first, filtering invalid data and performing deduplication based on the existing feature data of patents; second, conducting missing data imputation and data standardization for each feature group respectively; third, annotating the data of each feature group. The core feature groups cover patent demand matching degree, patent classification matching degree, market willingness, grant timeliness, and patent text quality, among others. The patent value estimates derived from this trained model can be applied to enterprise patent portfolio analysis and patent value analysis.

搜集汇总

数据集介绍

特点

该数据集是一个基于大数据训练的专利价值计算模型，用于计算专利的估算价值，适用于企业专利布局和专利价值分析。数据集包含35200条数据，格式为csv。