A dynamic entity graph is similar to a named entity graph, with the difference that we can define it dynamically through the Java API. The following is an example of how to define an entity graph using the Java API:

EntityGraph<?> entityGraph = getEntityManager().createEntityGraph(Account.class);
entityGraph.addSubgraph("transactions");
Map<String, Object> hints = new HashMap<String, Object>();
hints.put("javax.persistence.fetchgraph", entityGraph);

return this.getEntityManager().find(Account.class, accountId, hints);

So, if we have lots of use case-specific entity graphs, this approach will be an advantage over named entity graph where adding an annotation on our entity for each use case makes code unreadable. We can keep all use case-specific entity graphs in our business logic. With this approach, the disadvantage is that we need write more code and in order to make code reusable, we need to write more of the methods for each related business logic.