Benefits of a Graph Query Language

[wassfila]

• When do we need a Graph Query Language ?
• When do we not need a Graph Query Language ?
• What are the alternatives ?
• What does this need depend on ?

Short answer

• GQL helps writing simple declarative queries for search and filtering purposes that would be too complex to write in a imperative form.

Graph Query Languages

https://www.gqlstandards.org/existing-languages

Graph Modelling Language

• GML : https://en.wikipedia.org/wiki/Graph_Modelling_Language
• Usage : Although being referred to as Modelling Language, the more appropriate context is file format and data exchange protocol.

GML with free Queries

• Given that the response format can be defined with the GML, the query can become freely defined as a way to filter content of the returned data
• search indexing becomes a main topic to improve query results efficiency

Analysis

• A GQL can be considered more of an analysis tool than a query tool
• Particularly important in complex data where relationships can be very deep

Assumption

• Shallow Graphs and mainly Tree oriented Graphs would not have a big advantage from using a GQL and can be processed with simple search filters
• Complex Graphs with deep relationships requiring a specific analysis would need a GQL to better filter the search results based on the graph structure
• Counter examples of complex graphs
  • almost all nodes are directly connected to all others
  • almost all nodes are indirectly connected to all others with few edges

[wassfila]

data exist in applications within different isolated clusters. Every data knows which other data it can interact with. Not havin a graph abstraction layer require to keep rewriting code specific to the data although the query is always the same. Therefore converting the data to a graph representation once can allow to use all available graph analysis frameworks.

Cypher, as a graph representation even, is more compact than a json nodes and edges list, even compared to packed json data. Cypher is a declarative language that is easy to comprehend given the direct descriptions of properties and relations.
The main question is, within a huge graph, how to easily filter data based on props and relations in order to get a smaller graph to visualize ? G.Q.L answers exactly that question.
Data that is not not stored in a graph format will always require coding overhead to process it. G.Q.L is therefore needed to easily analyse any data in a generic abstracted way.
note existing languages
https://www.gqlstandards.org/existing-languages