J-Ben's main window has two tabs, providing a bidirectional word dictionary and a kanji (character) dictionary. These dictionaries are dependent upon the various dictionary files (EDICT2, KANJIDIC/KANJD212/KANJIDIC2, KRADFILE/RADKFILE) freely available from Monash University on Jim Breen's homepage.
The word dictionary performs a very simple search based upon the ENTIRE QUERY - it doesn't try to match all the words or some of the words, but rather treats your query as a phrase and matches only the whole phrase. Results are arranged with exact matches first, followed by "begins with" matches, "ends with" matches, and finally, everything else. Japanese searches work best because of the structure of EDICT2, but English searches work fairly well as well.
The word dictionary ties into the vocab study list. You can step through or jump to a random word in your vocab list using the buttons at the bottom of the dictionary tab.
The kanji dictionary searches for one or more kanji characters and displays detailed information about each one. Because of the amount of data available for each character, and because most people don't need much of the data available, the output can be customized through the preferences editor.
Information provided includes (but is not limited to) the following:
Display of stroke order diagrams can be toggled here. This is dependent upon installing a set of stroke order diagrams.
The kanji dictionary ties into both the kanji and vocab study lists. You can move through your kanji study list using the forward, backward, and random buttons at the bottom of the dictionary tab. Further, if you search for a single kanji and it is in your list, the dictionary will recognize this and allow you to step through your list from the current character. Finally, all character searches are cross-referenced with your vocab study list, and if any of your vocab contain the kanji, they will be listed as well.