• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Conference on Data Engineering and Communication Technology >Explore Web Search Experience for Online Query Grouping and Recommendation by Applying Collaborative Re-ranking Algorithm
【24h】

Explore Web Search Experience for Online Query Grouping and Recommendation by Applying Collaborative Re-ranking Algorithm

机译：通过应用协作重新排名算法探索网上查询分组和推荐的网络搜索体验

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Millions of users are Web dependent to search relevant information of complex tasks. Query clustering is a collection of relevant queries which are previously searched and related to currently issued query. A complex task is divided into a number of smaller tasks, and multiple queries are searched for each task repeatedly. Searching task-related information online is still textually based. But it suffers from the problem of polysemy and synonymy queries. K-Means algorithm solves problem of textual similarity, but there is no flexibility to increase the number of clusters. Graph-based query clustering method is used to detect similarity between current query and existing query group by exploring collaborative search history. Online clustering algorithm provides facility to create query group dynamically. Collaborative re-ranking algorithm improves search performance by recommending highly relevant searched results by ranking queries in query group. Several experimental results indicate the proposed system has higher precision and recall values.
机译：数百万用户依赖于Web，以搜索复杂任务的相关信息。查询群集是先前搜索的相关查询的集合，并与当前发布的查询相关。复杂任务分为多个较小的任务，重复搜索每个任务的多个查询。在线搜索与任务相关的信息仍然是基于的。但它遭受了多义和同义词查询的问题。 K-means算法解决了文本相似之处的问题，但没有灵活地增加集群的数量。基于图形的查询群集方法用于通过探索协作搜索历史记录来检测当前查询和现有查询组之间的相似性。在线聚类算法提供了动态创建查询组的工具。协同重新排名算法通过在查询组中的查询中推荐高度相关的搜索结果来提高搜索性能。几个实验结果表明，所提出的系统具有更高的精度和召回值。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

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