The architecture for the project includes four main components:

• Data collection:
  • IGT detection (Xia and Lewis, IJCNLP-2008)
  • Language ID: (Xia et al., EACL-2009)
• System projection: (Xia and Lewis, NAACL-2007)
• Bootstrapping NLP tools: (Georgi, 2009)
• Cross-lingual study: (Lewis and Xia, IJCNLP-2008)

An overview of the project is described in (Xia and Lewis, LaTeCH-SHELT&R 2009).