【24h】

Prepositions and Conjunctions in a Natural Language Interfaces to Databases

机译:数据库的自然语言介词中的介词和介词

获取原文
获取原文并翻译 | 示例

摘要

This paper present the treatment of prepositions and conjunctions in natural language interfaces todatabases(NLIDB) that allows better translation of queries expressed in natural language into formal languages. Prepositions and conjunctions weren't sufficiently studied for their usage in NLIDBs, because most of the NLIDBs just look for keywords in the sentences and focus their analysis on nouns and verbs getting rid of auxiliary words in the query. This paper shows that prepositions and conjunctions can be represented as operations using formal set theory. Additionally, since prepositions and conjunctions keep their meaning in any context, their treatment is domain independent. In our experiments we used Spanish language. We validate our approach using two databases; Northwind and Pubs of SQL Server, with a corpus of 198 different queries for the first one and 70 queries for the second one. The 84% of queries were translated correctly for the database Northwind and 80% for Pubs.
机译:本文介绍了自然语言到数据库(NLIDB)的介词中的介词和连词的处理方式,从而可以更好地将以自然语言表达的查询转换为形式语言。介词和连词在NLIDB中的用法还没有得到充分的研究,因为大多数NLIDB只是在句子中寻找关键字,而将其分析重点放在名词和动词上,从而摆脱了查询中的辅助词。本文表明介词和连词可以用形式集理论表示为运算。另外,由于介词和连词在任何情况下都保持其含义,因此它们的处理是独立于域的。在我们的实验中,我们使用西班牙语。我们使用两个数据库来验证我们的方法; SQL Server的Northwind和Pubs,第一个查询具有198个不同查询的语料库,第二个查询具有70个查询的语料库。对于数据库Northwind,正确转换了84%的查询,对于Pubs,则正确转换了80%。

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