首页> 外国专利> RANKING FUNCTION GENERATING DEVICE, RANKING FUNCTION GENERATING METHOD, AND RANKING FUNCTION GENERATION PROGRAM

RANKING FUNCTION GENERATING DEVICE, RANKING FUNCTION GENERATING METHOD, AND RANKING FUNCTION GENERATION PROGRAM

机译:排名函数生成装置,排名函数生成方法以及排名函数生成程序

摘要

PROBLEM TO BE SOLVED: To generate a ranking function using features which cannot be acquired with respect to all the Web pages.;SOLUTION: A feature extraction function part 20 acquires storage information of a DB30 for each page identifier of a DB10, and calculates a feature value from the frequency of appearance of a search query in a text and the rank of the DB30, and stores it in a DB40. Click information of the page identifier, search query, and Web page are stored in a DB50. An estimator generation function part 60 generates an estimator having single variable function/multiple variable function showing a relation between the click information and the feature value for each page identifier and search query of the DB50. Estimation click information calculated by inputting the feature value of the feature database to the function of the estimator is stored in a DB80. A ranking function generation function part 90 performs ordered learning from the rank of the DB30, the feature value of the DB40 and the estimation click information of the DB80, and generates the ranking function.;COPYRIGHT: (C)2011,JPO&INPIT
机译:解决的问题:为了使用不能针对所有网页获取的特征来生成排名函数。;解决方案:特征提取功能部分20针对DB10的每个页面标识符获取DB30的存储信息,并计算根据文本中搜索查询出现的频率和DB30的排名确定特征值,并将其存储在DB40中。页面标识符,搜索查询和网页的点击信息存储在DB50中。估计器生成功能部分60生成具有单变量函数/多变量函数的估计器,该单变量函数/多变量函数示出针对DB50的每个页面标识符和搜索查询的点击信息与特征值之间的关系。通过将特征数据库的特征值输入到估计器的功能而计算出的估计点击信息被存储在DB80中。排名函数生成功能部90根据DB30的排名,DB40的特征值和DB80的估计点击信息进行有序学习,并生成排名函数。COPYRIGHT:(C)2011,JPO&INPIT

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