My question is motivated by the following problem. I have a set of web documents from which I extract keywords. I want to store these data in Neo4j for further analysis (more or less graph mining including subgraph isomorphism problem): each web document is a node; hyperlink from one web document to another is a corresponding directed relationship; keywords are properties of the nodes. In this setting, keyword property may be attributed to several nodes (I hope this is doable).
I need help with the following questions (which I find quite difficult to answer knowing only very basic things about Neo4j):
1) Is it possible to select all nodes attributed with a specific property "keyword1"?
2) How can I select common (overlapping) keyword properties for 2 nodes "doc1" and "doc2"? i.e., common keywords for 2 web documents
3) Is it better to create some kind of string key for keyword properties (rather then use default auto-incremented integer)?
Any hints/recommendations/links will be highly appreciated. I am using Python binding for Neo4j on Windows.
