Question

我可以编写一个SPARQL查询来将单个三元组中的单词中的单词连接成单个句子：

SELECT  ?sent   (sample (?sf1) as ?sf11) (group_concat(?wf)  as ?sf2)  
WHERE 
{   ?sent a nlp:Sentence .
    {select *
    { ?w rdfs:partOf ?sent . 
      ?w a nlp:Token .
      ?w nlp:wordForm ?wf . 
    }   order by ?w 
  }
}  group by ?sent   
  limit 20

但是我找不到在另一个查询中使用这个select语句的方法，在那里我找到句子?sent并想插入这个select语句：

select * 
where  
{   ?tok a wn:Locomote. 
    ?tok nlp:lemma3 ?lem .
    ?tok rdfs:partOf ?sent .
    ?sent a nlp:Sentence .     
    ?werk ^rdfs:partOf ?sent . 
    {SELECT ?sent  (group_concat(?wf)  as ?sf2)  
    WHERE 
    {   ?sent a nlp:Sentence .
        {select *
        { ?w rdfs:partOf ?sent . 
          ?w a nlp:Token .
          ?w nlp:wordForm ?wf . 
        }   order by ?w 
      }
    }  group by ?sent  
  }
}

结果不是第一部分中找到的句子，但似乎嵌套查询中的?sent不受外部查询的限制。我没有看到如何正确嵌套。谢谢你的帮助！

Answer 1

Example of custom aggregate in Jena:

    <TestMethod()> Public Sub ElementTypeUnitTestAsync()
       Dim RoslynPath As String = Path.Combine(RepoPath, SolutionPartialPath)

        Dim MS_Workspace As MSBuildWorkspace = MSBuildWorkspace.Create()
        AddHandler MS_Workspace.WorkspaceFailed, Sub(sender As Object, e As WorkspaceDiagnosticEventArgs)
                                                     Debug.WriteLine(e.Diagnostic.ToString())
                                                 End Sub
        Dim NewSolution As Solution = MS_Workspace.OpenSolutionAsync(RoslynPath).Result
    End Sub

使用sparql查询作为另一个查询中的构建块

1 个答案: