Finding trends in loads of text data

Question

Hi:

I have a large table and would like to at least find the most repeated works so I can have a trends keyword list.

The database is mysql and I am running php5.

Is there any other way besides exploding the data and searching for repetitions on the php side?

What about having combinations of 2/3 words keyword trends?

Any thoughts?

Thanks

Answer 1

well... as I didnt get any answers I found a way using php:

$wordfrequency = array_count_values( str_word_count( $string, 1) );

hope it helps someone :)

Answer 2

It sounds like you are almost talking about implementing a Markov chain sort of algorithm. Without processing the data fully, though, I cannot think of any way you could use SQL alone to do that sort of trending.