The reason for that is that people who have capabilities of doing that know that its not a good thing.
The data can be used for re identification attacks against user privacy.
Some good advice
1. use python + NLTK
2. read papers and posters from current conferecnce such as ACL or even WWW
3. start an incremental project and try at least getting a poster published in a tier-1 conference
4. If you have good grades + work ex go to CMU Advanced Lang. Tech Masters program
At the same time you must look out for IQ and Looks which are quite a bit genetically dependant