I am working with a big indexing system, that will index almost 10 years of a newspaper,
about 300 articles/day, about 4 Kb (350 words) each : 10*365*300 texts = 1.095.000 of
articles (about 4,3 Gb of Data).
It's my first use of the Full Text Search of MySQL, and I am having some problems:
1st) I like to acess the Dictionary Table directly, to retrieve all the world that was
indexed. This will be used to construct the front end, allowing the user only to make a
search based on world that really exists in the dictionary table.
2nd) I like to limit the grown of the dictionary to words from 4 letters to 25 letters.
3rd) How I can grown the search speed ? The primary search is allways COUNT(*) of
ocorrences of some word in the text table.
4rd) I made some tests of performance, and I noted that the first search for a very large
occurence (abour 1% of the data) takes 100 times the secound iqual search, this smell to
me a very sofisticated cache mechanism: How it works ? Could I have some information on
that ? I could apprimorate it ?