英文:
In which case should we use Neo4j or other graph database?
问题
我想制作一个博客网站，我需要一个评论系统（用户可以为每篇文章添加评论）。
考虑以下因素：
- 保存大量评论
- 我会保存点赞数量
- 我需要对这些评论进行排名（评论的分数由点赞数量和时间计算，随着时间的推移会发生变化）
由于将有大量的插入（添加评论）和更新（更新点赞数量）操作，而且我需要根据评论的分数来选择评论，我认为关系型数据库不适用。
Neo4j适合这种情况吗？
我想将每篇文章和评论都保存为Neo4j的节点。
英文:
I want to make a blog website and I need a comment system (user can add comments for each article).
take those into consideration:
- save lots of comments
- I will save like count
- I need to rank those comments (the score of a comment is calculated by like count and time, and will change with time)
As there will be lots of insert(add comments) and update(update like count) operations, and I need to range select comments by score of comment, I think relational database is not suitable.
Is Neo4j suitable for this situation?
I want to save each articale and comment as a node of neo4j
答案1
得分: 1
> 由于将有大量的插入（添加评论）和更新（更新喜欢计数）操作，以及我需要按评论分数范围选择评论，我认为关系数据库不适用。
这很有趣，因为这基本上是关系数据库擅长的事情之一。
话虽如此，这是一个相当简单的任务，任何数据库，无论是关系型还是其他类型的，都应该能够处理，直到评论数量达到了数百万。
> 那么，Neo4J是否适用于这种情况？
如上所述，我相信Neo4J可以处理。但我没有看到任何与图数据库的优势相匹配的内容。您的图仅包含两种类型的节点（article和comment），每个comment引用了一个article。
我个人认为，在平面文件不再足够的情况下，关系型数据库管理系统（RDBMS）是默认选择。所以我会在这里选择RDBMS。但再次强调，您提供的要求应该很容易由任何数据库处理。
英文:
> As there will be lots of insert(add comments) and update(update like count) operations, and i need to range select comments by score of comment, I think relational database is not suitable.
That is interesting, because it is pretty much one of the things a relational database excels at.
That said it is a rather trivial tasks and any database relational or otherwise should be able to handle this until you get into many millions of comments.
> so, is neo4j suitable for this situation?
As said above, I'm sure Neo4J can handle it. But I don't see anything that plays to the strengths of a graph database. You don't have a complex graph. Your graph simply consists of two types of nodes (article and comment) with each comment referencing exactly one article.
I personally consider an RDBMS as the default choice once a flat file isn't sufficient anymore. So I'd pick an RDBMS here as well. But again, the requirements you gave should be easy to handle by any database.
答案2
得分: 1
你所描述的并不太“图形化”。当出现以下情况时，它会更加“图形化”：
- 想要知道哪些用户在他们评论的事物上相似
- 拥有一个层次化系统来“标记”帖子和评论
- 拥有具有任意深度的分层评论
英文:
what you describe is not very "graphy". It becomes more graphy when:
- you want to know which users are similar with respect to the things they comment on
- you have a hierarchical system to "tag" posts and comments
- you have threaded comments with arbitrary depth
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论