After looking for information about creating an in-memory dictionary search, i came to the conclusion that a tries are the best way to do this. The word trie comes from retrieval, so the algorithm is good for retreival of fixed text (Such as our dictionary). It is good in situations where the time in preprocessing the text is minimal considering it would be offset by very fast retreivals later on. I already have a webserver written during my .NET class . It is a simple multithreaded HTTP server and i think it could be adapted easily to create a dictionary server. This will eliminate the need for a database. This approach is only good for custom purposes like this dictionary project, but when developing a custom ASP.NET web control or something similar, we should stick to standard stuff such as connectivity to traditional database servers etc.