Dogpile Dictionary

This essay is about a suggested student project in Java programming. This essay gives a rough overview of how it might work. It does not describe an actual complete program. I have no source, object, specifications, file layouts or anything else useful to implementing this project. Everything I have to say to help you with this project is written below. I am not prepared to help you implement it; I have too many other projects of my own.

I do contract work for a living, which could include writing a program such as this. However, I don’t do people’s homework for them. That just robs them of an education.

You have my full permission to implement this project any way you please.

This project is similar technically to the Bookstore Referral Project and the commercial Dogpile search engine that works by asking several other search engines to do its work for it in parallel. Have a look at the impressive list of dictionaries indexed at YourDictionary.com . It contains references to multiple web-based dictionaries in over 200 languages so you can translate from any language into any other. It even has dictionaries for Klingon, Romulan, Vulcan and Tolkien. How do you manually make use of this wealth?

You make a list of the dictionaries that handle the particular translation you want, and then experiment with them to find which is most likely to have the words you want and the best quality definitions. You then try them in order.
If you still can’t find the meaning of a word, you try using an intermediate third language, attempting a two-step translation. For example, if you wanted to translate from Indonesian to Vulcan, you might do it via English or Esperanto .

Your project, should you decide to take it, is to automate this process. There are several tasks:

Find all the on-line dictionaries, and write code to scan them and convert the information into a standard format. The dictionary entries will be constantly changing and the dictionary formats too will be more slowly changing. You will thus likely look words up on-line rather than batch convert the dictionaries, though that two may be feasible, if you have a central server. For a start, deal only with dictionaries organised with the <dl><dt></dt><dd></dd></dl> HTML glossary tags.
Given the particular translation task, select a subset of dictionaries to use.
Based on the user’s past success, and the general population’s past success, rank the dictionaries so that you try the best first. You might consider using parallel tasks to help you, potentially running on many different servers.
If that search fails, try an intermediate language. You get a combinatorial explosion here, so I don’t think using a 3 step process with two intermediate languages would be feasible.
Package this up much like the Bookstore Referral Applet. The code must be frequently updated, so ensuring users get the freshest code is quite important. Unfortunately Applets may not talk to multiple foreign servers, so it may be simplest to make this an application.
Maintain a local cache of words already looked up so you can find them quickly without going to the Internet.
Devise code to peel off standard suffixes and prefixes when you can’t find a word, to get at the root which may be in one of the dictionaries.

I debated whether to warn you that this project, done properly, is trickier than it first looks. Just the same, it needs to be done, even if done badly. Working prototypes may encourage more serious attempts. I am working on this myself. I have written a prototype for Esperanto-English that at present uses but a single dictionary. Have a look at my Esperanta Tradukilo Vortope.

	Please email your feedback for publication, errors, omissions, broken/redirected link reports and suggestions to improve this page to Roedy Green :
	Canadian Mind Products
	mindprod.com IP:[65.110.21.43]
	Your face IP:[38.103.63.18]
	You are visitor number 3,717.
	You can get a fresh copy of this page from:	or possibly from your local J: drive (Java virtual drive/Mindprod website mirror)
	http://mindprod.com/project/dogpiledictionary.html	J:\mindprod\project\dogpiledictionary.html