如何通过PHP使用SPARQL向Fuseki插入数据并在网页实现类Protege RDF图谱

首先需要确认Fuseki对应数据集的更新端点（默认格式为http://你的Fuseki地址:3030/数据集名称/update）已开启写权限，无需额外依赖PHP扩展，直接通过cURL发送SPARQL更新请求即可，示例代码如下：

<?php
// 替换为你的Fuseki更新端点地址
$fusekiUpdateUrl = 'http://localhost:3030/dl_lib/update';
// SPARQL插入语句，可直接匹配你提供的本体结构
$sparql = <<<SPARQL
PREFIX dl: <http://www.semanticweb.org/aungyekyaw/ontologies/2020/8/DL_lib#>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
INSERT DATA {
  dl:Book_3 rdf:type dl:Book ;
            dl:hasAuthor dl:Ivan_Marsic ;
            dl:isInCategory dl:Computer_Science,_Information_and_General_Works ;
            dl:description "As the author of computer books, I spend a lot of time lurking in the computer section of bookstores, observing the behavior of readers while I’m pretending to read the latest issue of Soap Opera Digest magazine.Because of my research, I’ve learned that if you have picked up this book and turned to this introduction, I only have 13 more seconds before you put it down and head to the coffee bar for a double-tall-decaf-skim-with-two-shots-of-vanilla-hold-the-whip latte." ;
            dl:hasISBN "ISBN-978-0-672-32990-6" ;
            dl:hasTitle "Software Engineering" ;
            dl:locatedIn "HEC" ;
            dl:published "2012-01-01T00:00:00Z"^^xsd:dateTime .
}
SPARQL;

$ch = curl_init($fusekiUpdateUrl);
curl_setopt($ch, CURLOPT_POST, true);
curl_setopt($ch, CURLOPT_POSTFIELDS, $sparql);
curl_setopt($ch, CURLOPT_HTTPHEADER, [
    'Content-Type: application/sparql-update',
    // 如果Fuseki开启了账号认证，取消注释下行替换为实际账号密码
    // 'Authorization: Basic ' . base64_encode('admin:你的密码')
]);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
$response = curl_exec($ch);
$httpCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);
curl_close($ch);

if ($httpCode == 200 || $httpCode == 201) {
    echo "数据插入成功";
} else {
    echo "插入失败，错误码：" . $httpCode . "，响应内容：" . $response;
}
?>

注意事项：

• 插入的日期类型属性需要符合XMLSchema的dateTime格式，避免Fuseki校验报错
• 若生产环境使用建议对SPARQL语句做参数化处理，避免注入风险

你提供的Protege导出的OWL个体示例如下：

<owl:NamedIndividual rdf:about="http://www.semanticweb.org/aungyekyaw/ontologies/2020/8/DL_lib#Book_3">
    <rdf:type rdf:resource="http://www.semanticweb.org/aungyekyaw/ontologies/2020/8/DL_lib#Book"/>
    <hasAuthor rdf:resource="http://www.semanticweb.org/aungyekyaw/ontologies/2020/8/DL_lib#Ivan_Marsic"/>
    <isInCategory rdf:resource="http://www.semanticweb.org/aungyekyaw/ontologies/2020/8/DL_lib#Computer_Science,_Information_and_General_Works"/>
    <description rdf:datatype="http://www.w3.org/2000/01/rdf-schema#Literal">As the author of computer books, I spend a lot of time lurking in the computer section of bookstores, observing the behavior of readers while I’m pretending to read the latest issue of Soap Opera Digest magazine.Because of my research, I’ve learned that if you have picked up this book and turned to this introduction, I only have 13 more seconds before you put it down and head to the coffee bar for a double-tall-decaf-skim-with-two-shots-of-vanilla-hold-the-whip latte.</description>
    <hasISBN rdf:datatype="http://www.w3.org/2000/01/rdf-schema#Literal">ISBN-978-0-672-32990-6</hasISBN>
    <hasTitle rdf:datatype="http://www.w3.org/2001/XMLSchema#string">Software Engineering</hasTitle>
    <locatedIn rdf:datatype="http://www.w3.org/2001/XMLSchema#string">HEC</locatedIn>
    <published rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2012</published>
</owl:NamedIndividual>

具体实现步骤如下：

• 第一步：导出/获取RDF数据，从Protege中将完整本体导出为JSON-LD、TTL或RDF/XML格式，优先选择JSON-LD，前端处理成本更低；也可以直接将数据导入Fuseki后，通过SPARQL查询动态获取最新数据。
• 第二步：数据格式转换，把RDF三元组转换为前端可视化库通用的结构：
  • nodes数组：每个元素对应图谱中的一个节点，包含唯一ID、显示标签、节点类型等字段，比如Book_3节点的标签是Software Engineering，类型为Book，还可以把ISBN、出版时间等属性作为扩展字段存在节点对象里。
  • edges数组：每个元素对应两个节点之间的关系，包含源节点ID、目标节点ID、关系标签，比如Book_3和Ivan_Marsic之间的边标签为hasAuthor。
• 第三步：选择前端可视化库，常用的有ECharts关系图、AntV G6、D3.js力导向图，都是纯前端实现，不需要依赖外部服务，直接引入即可使用。
• 第四步：渲染与自定义配置，把转换好的nodes和edges传入可视化库的配置项，可自定义节点颜色、大小，边的样式，点击节点展示属性详情等交互，最终实现和Protege中类似的图谱展示效果。

内容的提问来源于stack exchange，提问作者Ye Kyaw