Question

Is it possible to find repeating patterns in the text?

My table looks like this:

CREATE TABLE `textanalysis` (
    `ID` int(11) NOT NULL AUTO_INCREMENT,
    `abstract` text,
    UNIQUE KEY `ID` (`ID`),
    FULLTEXT KEY `abstract` (`abstract`)
) ENGINE=MyISAM AUTO_INCREMENT=2 DEFAULT CHARSET=latin1;

I would like to find the words or group of words in the text then make a statistics.

Answer 1

Here is some tricks (not very optimized)

use “apple” for example,
length for apple is 5

SELECT
  (LENGTH(abstract)-LENGTH(REPLACE(LOWER(abstract), 'apple', '')))/5 
  AS occurrences
FROM
  textanalysis
WHERE
  MATCH (abstract) AGAINST ('+apple' IN BOOLEAN MODE);

What is does is to replace apple (make the length of abstract shorter),
and you compare the original length to deduce number of occurrences.