Facebook Search: Information Search Algorithm

Question

There is a closed-ended question titled “ How does the search feature on Facebook work?

In simple terms, the OP asked (and even gave a sample of what he was trying):

How does Facebook Graph Search work? He gave an example: Friends from France who likes England

How can this be realized as a problem of finding information in the real world?

As my answer does not fit in the comment, so I thought about re-asking the question and answering it well in the style of Qaru Q & A.

Answer 1

In the perspective of implementation, pay attention to the property graph, for example, on Neos4j and Lucene, based on NoSQL, as a search engine for the graph database (+ Distributed Systems). The example provided by you, "Friends from France who love Russia," can be easily handled by Neo4j or a graphical solution of an arbitrary distributed property.

Suppose you use Neo4j, in this case, France will be the value (in the node graph) in the key value store, where say key = Country, all outgoing edges from this node with a “like” label should be traversed, and all the end vertices will be again look for England. What is it (of course, the task here would be to Search and Workarounds for Distributed Degrees of Big Data)!

Consider a diagrammatic view taken from the Neo4j site for a property graph:

I still haven't read about searching on Facebook though :)