The Tfitter project aims to store Twitter data with Scala in backends
such as 

-- PostgreSQL
-- Berkeley DB Java Edition
-- Lucene
 
for subsequent analysis of the contents and the conversations 
in the context of Machine Learning and Natural Language Processing.

The originator, Alexy Khrabrov, is a researcher at Thayer School of Engineering at Dartmouth.

http://www.khrabrov.net/ points to Alexy's personal and academic pages.
See Alexy's Scala blog,

http://la.scala.la/

for Scala discoveries in the making of Tfitter.