JMdictDB - Japanese Dictionary Database

Editing Quickstart

Search | Advanced Search | New Entry | Submissions | Help
Login for registered editors

Getting Started

This is a short introduction to submitting corrections to the JMdict/Edict dictionaries. For full details of the syntax required please see the full JEL reference or the "help" links on the New Entry and Edit Entry forms. Please also be sure you are familiar with the JMdict/EDICT Editorial Policy.

Entries in the JMdictDB database are usually in three parts: a kanji/special character part, a reading(s) part (always in kana) and a meaning(s) part. Consider the following entry as it appears in WWWJDIC:

思い出す 【おもいだす】 (v5s,vt) to recall; to remember

Here 思い出す is the kanji/special character part, おもいだす the reading(s) part and "(v5s,vt) to recall; to remember" the meaning(s) part.

To enter a simple entry such as the one above, key or paste the information of the parts in the three boxes. Put 思い出す in the kanji/special character part, おもいだす in the readings part, and in the meanings part put:

[1][v5s,vt] to recall; to remember

That meanings format needs a little explanation:

  1. the "[1]" is the number of the "sense", i.e. a related group of meanings. All entries have at least one sense.
  2. the "[v5s,vt]" is the coded "part-of-speech", in this case saying it's a "Godan" verb (五段) ending in す (サ行), and it's transitive. (The full set of part-of-speech tags can be seen by clicking on the link to the right of the box.)
  3. the two English meanings are simply separated by semicolons. (If you want a semicolon inside a meaning, use "\;" to tell the system it is not a separator.)
In some cases where the word/expression is written in kana alone, the kanji/special character part is omitted, e.g. in
アイスクリーム (n) ice cream; icecream

Be sure to mention the source of your entry (name of dictionary, URL of WWW site, etc.) in the References box, and feel free to comment on the entry or your submission in the Comments box. (Note that the contents of these two boxes are kept in the database, and can be viewed by others as they form part of the history of the entry.)

Once you have completed the entry (or edit), click on the "Next" button at the foot of the screen. If no errors are detected, you will be shown what the entry will look like in a more readable form. If you are happy with it, click on "Submit", otherwise go back and amend it.

Digging a Bit Deeper

The information above should be enough to get you started, but let's just look at a couple more points.

1. That clumsily named "kanji/special character part". This for the form of the entry that is written using at least one kanji character or other non-kana character, e.g. 漢字, DNA, 〒, 2ちゃんねら, etc. would all go in this part as each has at least one non-kana character. (In linguistics, this is often called the "surface form" of a word.)

2. Quite a few entries have multiple versions of the kanji part or the readings part, e.g.:

  一結び; ひと結び 【ひとむすび】 (n,vs) knotting
  金肥 【きんぴ; かねごえ】 (n) chemical fertilizer
You can enter these multiple versions simply by separating them with a ";", just as in the extended EDICT format above: 一結び; ひと結び in the Kanji box or きんぴ; かねごえ in the Readings box.

3. Each of the parts of the dictionary entry can be accompanied by a number of tags, such as the part-of-speech mentioned above. In this system the tags are written in a simple language (read about it in the full Help File) and placed inside square backets: [..]. Most of these tags are used as-is and can be used in any order. They can also be grouped; "[n,uk,abbr]" is the same as "[n][uk][abbr]". Some, such as cross-references, need a value, e.g. [see=時々] or [note="contraction of 書いておく"]. (If you want to use square brackets within a meaning, type \[ and \] to tell the system they are not tags.)

4. The meanings part of the entry will contain or more "senses", i.e. different sets of meanings for the term, as in:

識別名 【しきべつめい】 (n) (1) {comp} identifier; (2) {comp} distinguished name
Use "[1]", "[2]", etc. at the start of each sense. So the above senses would be entered as:
[1][n,comp] identifier [2][n,comp] distinguished name
5. Note that the ";" is used for separating kanji forms, readings snd glosses within a sense. Don't use them at the end of the text in a box, or after the last gloss in a sense. That will cause an error.

And In Closing

The system using [...] to enter tags, notes. cross-references, etc. avoids having to have forms with masses of specific boxes. It can, however, take some time to get used to it.

If you find it too complicated, write a summary of what you want in the "Comments" box and an editor will fix up the format.