首页> 外国专利> THE PROCEDURE FOR GROUPING PATENT DOCUMENTS BASED ON INTERNATIONAL CLASSIFICATION CODES AND TECHNOLOGY GROUPS

THE PROCEDURE FOR GROUPING PATENT DOCUMENTS BASED ON INTERNATIONAL CLASSIFICATION CODES AND TECHNOLOGY GROUPS

机译:基于国际分类代码和技术组的专利文件分组程序

摘要

A method for the classification of patent documents by using international classification codes and product groups and manufacturers has for a novelty the specific way of calculating the coefficient of belonging and automatic withdrawal of the mapped products and manufacturers that are corresponding to the patent documents that have to be classified. The method comprises phase (100) including extraction of classification code from the patent documents and their automatic grouping to the groups of products and manufacturers. The method further comprises phase (101) in which the matrix is formed including: patent numbers, classification codes and groupes of products and manufacturers. The method includes also phase (103) in which coefficients of belonging of each classification codes within the groups (200) are calculated. The coefficient of (301) belonging depends of frequency of occurrence (203) of classification code (202) within the specified group (200), the total number of classification codess and the the total number of patent documents within one group. Phase (104) comprises forming of a matrix of classification of patent documents within mentioned groups, where is for each patent document in the list (201) of patent document determined coefficient (301) of belonging to each group (200), respectively. In phase (105) the algorithm determines the maximum (300) of coefficient of belonging. The proposed phases of method are executing in the classification unit (405) which communicates via communication network (402) with user unit (400) and the servers (403) which contain data related to products and manufacturers and also it communicates with servers (408) with available patent databases.
机译:一种使用国际分类代码和产品组以及制造商对专利文件进行分类的方法,其新颖之处在于,它有一种特定的方式来计算与必须与专利文件相对应的被映射产品和制造商的归属系数和自动撤回。被分类。该方法包括阶段(100),该阶段包括从专利文件中提取分类代码并将其自动分组到产品和制造商的组。该方法还包括阶段(101),其中形成矩阵,该矩阵包括:专利号,分类代码以及产品和制造商的组。该方法还包括阶段(103),其中计算组(200)内的每个分类代码的归属系数。属于的系数(301)取决于指定组(200)内分类代码(202)的出现频率(203),分类代码的总数和一组内的专利文件的总数。阶段(104)包括形成所述组内的专利文件的分类矩阵,其中分别针对专利文件列表(201)中的每个专利文件确定属于每个组(200)的系数(301)。在阶段(105)中,算法确定归属系数的最大值(300)。方法的建议阶段在分类单元(405)中执行,该分类单元(405)通过通信网络(402)与用户单元(400)和服务器(403)通信,其中用户单元(400)包含与产品和制造商有关的数据,并且还与服务器(408)通信)和可用的专利数据库。

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号