首页> 外文会议>Multilingual information access in South Asian languages >Query Expansion Based on Equi-Width and Equi-Frequency Partition
【24h】

Query Expansion Based on Equi-Width and Equi-Frequency Partition

机译:基于等宽和等频分区的查询扩展

获取原文
获取原文并翻译 | 示例

摘要

Query Expansion has been widely used to improve the effectiveness of conceptual search. In this paper pseudo relevance feedback is used along with equi-width and equi-frequency partition technique. The proposed method effectively uses the position and frequency of the query terms for identifying a region within the retrieved documents, which is expected to contain expansion terms. This region is an intersecting region obtained by partitioning the retrieved documents using equi-width and equi-frequency partition techniques. Initial results indicate that words falling in the intersecting region contain good candidate terms for query expansion. The experiments are performed on FIRE 2011's Ad-hoc Hindi and English Data using Terrier as the retrieval engine. The initial experiments show an improvement in average precision of 12-14% in case of English data and 12.75% in case of Hindi data set.
机译:查询扩展已被广泛用于提高概念搜索的效率。在本文中,伪相关反馈与等宽度和等频率划分技术一起使用。所提出的方法有效地利用查询词的位置和频率来识别检索到的文档中的预期包含扩展词的区域。该区域是通过使用等宽和等频分区技术对检索到的文档进行分区而获得的相交区域。初步结果表明,落在相交区域中的单词包含用于查询扩展的良好候选词。实验是在FIRE 2011的临时印地语和英语数据上进行的,并使用Terrier作为检索引擎。初始实验显示,对于英语数据,平均精度提高了12-14%,对于印地语数据集,则提高了12.75%。

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