Dictionary with trie tree Java

Dictionary.java

//Dictionary implemented using a Trie Tree.
public class Dictionary {
	private HashMap<Character,Node> roots = new HashMap<Character,Node>();
	
	/**
	 * Search through the dictionary for a word.
	 * @param string The word to search for.
	 * @return Whether or not the word exists in the dictionary.
	 */
	public boolean search(String string) {
		if (roots.containsKey(string.charAt(0))) {
			if (string.length()==1 && roots.get(string.charAt(0)).endOfWord) {
				return true;
			}
			return searchFor(string.substring(1),roots.get(string.charAt(0)));
		} else {
			return false;
		}	
	}
	
	/**
	 * Insert a word into the dictionary.
	 * @param string The word to insert.
	 */
	public void insert(String string) {
		if (!roots.containsKey(string.charAt(0))) {
			roots.put(string.charAt(0), new Node());
		}
		
		insertWord(string.substring(1),roots.get(string.charAt(0)));
	}
	
	//Recursive method that inserts a new word into the trie tree.
	private void insertWord(String string, Node node) {
		final Node nextChild;
		if (node.children.containsKey(string.charAt(0))) {
			nextChild = node.children.get(string.charAt(0));
		} else {
			nextChild = new Node();
			node.children.put(string.charAt(0), nextChild);
		}
				
		if (string.length() == 1) {
			nextChild.endOfWord = true;
			return;
		} else {
			insertWord(string.substring(1),nextChild);
		}
	}
	

	//Recursive method that searches through the Trie Tree to find the value.
	private boolean searchFor(String string, Node node) {
		if (string.length()==0) {
			if (node.endOfWord) {
				return true;
			} else {
				return false;
			}
		}
		
		if (node.children.containsKey(string.charAt(0))) {
			return searchFor(string.substring(1),node.children.get(string.charAt(0)));
		} else {
			return false;
		}
	}
}

Node.java

public class Node {
	public Node parent;
	public Boolean endOfWord = false; //Does this Node mark the end of a particular word?
	public HashMap<Character,Node> children = new HashMap<Character,Node>();
}

What is the purpose of parent variable ?

-Node should be private inner-class so as not to expose internal data, or at least have private fields with package exposure
-Should remove useless last if-else branches since there isn't any code afterwards and just return false at the end of methods
-Use Boolean Zen when returning true/false based on node.endOfWord, i.e. just return node.endOfWord

@mithuns my guess is the parent field of nodes is there for future reverse searches of words.

Whats the time complexity of this?

this code is genius. love it! - I'd like to use parts of your logic in an application that I'm making, would that be okay with you?

I prefer TreeMap to store word than HashMap, it automatically sorts all the character

Here's a kotlin implementation spun off of this one: https://gist.github.com/bseib/2c2d2d4745c47353a884a51af0cc4a7d