The overall objective of NITE is to build an integrated best practice workbench for multi-level, cross-level and cross-modality annotation, retrieval and exploitation of multi-party natural interactive human-human and human-machine dialogue data.
Building on the MATE workbench and on the Noldus Observer software, the project aims at:
- specifying the architecture, the functionality and the usability of the NITE workbench to support best practice handling of the speech, the gesture and the facial expression;
- defining best practice coding schemes for the gesture, the facial expression and the cross-modality coding;
- specifying a markup framework;
- constructing a workbench according to the specification and incorporating coding schemes for the gesture, the facial expression and cross-modality issues;
- documenting the NITE workbench thoroughly;
- ensuring its usability for both expert and (computationally) novice users communities through an extensive testing with differently skilled users;
- evaluating the workbench iteratively with world-wide users to make it a standard.