Basic conception for database

[Rafal Drezek]

Hello,

I have academic question regarding DB performance. I am not DB expert but have some basic understanding about DB's and how to build simple SQL queries.

I want to build online database (PHP + MYSQL) providing some translations. Here I expect high load.

The questions is what will be the best approach to organize a tables.
Lets take this example:
Table A contains words in one language (with unique ID field)
Table B contains words in another language (with unique ID field)
Table C contains matching (connects words from table A and B considering the fact that one word from A can have many matches from B and oposite way).

If somebody wants to search the DB I can do 3 statements:
(1) select all matching fields from A
(2) for or results from 1 go to tha matching table C and find what would be translation from B
(3) read translations from C

OK. I know its not good approach, becuase one join can solve it. Right?
Can anybody provide me a good example for JOIN for given case? (assuming that tables A and B contain ID + word, table C contains two fields: ID from A, ID from B)

Another possibility:
What if I create one table (D) with: ID , word in one language, word in second language.
Here is one problem: I will have several same entries in one field (one word can have several translations), but even if for the previous case I build join, shouldn't it be better (regarding performance).

My goal is to have best performance.

Thanks for any suggestions on that.

regards
R.