Java : Datastructure to stock lots of words

Question

I have to stock lots of word (+200k) in a Java program and I want to access them really fast. I just need to know if a given word belongs to my "dictionary". I don't need a pair like <word, smthg>. If possible I'm searching a solution in the standard library.

PS : Maybe using a data structure is not the better way to do this ? Reading each time the file containing the words will be more efficient ?

edit : It's a small project. I have to deal with effectiveness and the memory

Last Edit : I finally choose HashSet.

Answer 1

Use java Sets because sets are linear sorted data structure like TreeSet. So for searching, techniques like binary search can be implemented and they are fast with no repetition.

This is the structure of a java Sets.

Also it will not going to allow duplication hence reducing redundancy and will save your memory.

If you want to know various searching algorithms complexities refer this link. Here is

http://bigocheatsheet.com/

Answer 2

Use either a Trie or Patricia tree depending on the distribution of the words. I would personally go with Patricia tree as it is more optimized for memory usage(though it is harder to implement).