七款国产知识图谱调研横向对比报告
郝伟 2021/03/14
调研报告下载地址
[
下载链接
]
附:相关图描述语言
本体语言学习笔记,
https://blog.csdn.net/shendeguang/article/details/8241164
OWL中基于具体化的多元关系表示研究,
https://doc.mbalib.com/view/1754a797116a45923bfaa7a1db86ce0e.html
知识图谱---描述语言,
https://blog.csdn.net/github_37002236/article/details/81908394
知识图谱基础之RDF,RDFS与OWL,
https://blog.csdn.net/u011801161/article/details/78833958
语义Web简单综述(XML、RDF、OWL、知识库、知识图谱),
https://blog.csdn.net/hohaizx/article/details/80043623
HugeGraph,Neo4j,Titan三种图数据库性能对比,
https://www.jianshu.com/p/c124f748877d
官网给了一个性能测试的报告:
https://hugegraph.github.io/hugegraph-doc/performance/hugegraph-benchmark-0.5.6.html
总结起来就是:
批量插入性能:HugeGraph(RocksDB) > Neo4j > Titan(thrift+Cassandra)
遍历性能:Neo4j > HugeGraph(RocksDB) > Titan(thrift+Cassandra)
图常用分析方法性能:FS场景,HugeGraph性能优于Neo4j和Titan,K-neighbor和K-out场景,HugeGraph能够实现在5度范围内秒级返回结果
社区聚类算法性能 Neo4j > HugeGraph > Titan