摘要
针对目前SPARQL(Simple Protocol and RDF(Resource Description Framework)Query Language)查询语句执行时间较长的问题,提出一种基于精英蚁群算法与权重矩阵的SPARQL查询语句优化算法。文中结合SPARQL针对不同图形状设计了有效的权重矩阵算法,因此本方法对不同查询形状均具有专门的优化参数;然后将权重矩阵作为蚁群算法的输入参数,分别利用人工蚁群与精英蚁群方法对SPARQL不同形状的查询语句进行优化。最终基于Apache Jena ARQ查询引擎进行实验,对比实验结果表明,该方法对于较宽广范围的查询语句均具有较好的优化效果,对三元组模式数量较多的查询语句效果尤佳,并且优化效果较为稳定。
Aimed at the problem that the execution time of the existing SPARQL query is too long, an elitistant algorithm and weight matrix based SPARQL query efficiency optimization algorithm is proposed. Anefficient weight matrix algorithm is designed with different shapes combined with SPARQL, so differentspecial optimization parameters are set to different query shapes; the weight matrix is used as the inputparameters of the ant colony algorithm, ant colony algorithm and elitist ant colony algorithm are used tooptimize the queries with different shapes. Lastly, experimental results based on Apache Jena ARQ queryengine show that the proposal gets good optimization results for wide range queries, especially for thequeries with more triple patterns, and the effect is more stable.
出处
《控制工程》
CSCD
北大核心
2017年第7期1439-1446,共8页
Control Engineering of China
关键词
精英蚁群算法
权重矩阵
查询语句
三元组模式
资源描述框架
Elitist ant algorithm
weight matrix
query statement
triple pattern
resource descriptionframework
