NITE (Natural Interactivity Tools Engineering) is a European HLT (Human Language Technologies) project which aims to develop a workbench, or an integrated toolset, for annotating and analysing natural interactive communicative behavior between humans, as well as between humans and machines (systems). NITE began its work in April 2001 and has a duration of two years. The NITE project partners are: NISLab (Odense, Denmark), DFE (Barcelona, Spain), DFKI (Saarbrücken, Germany), HCRC (Edinburgh, UK), ILC (Pisa, Italy), IMS (Stuttgart, Germany) and Noldus Information Technology (Wageningen, The Netherlands).

The vision of natural interactivity is that of enabling systems to exchange information with humans in the same ways in which humans exchange information with one another, i.e. through speech, facial expression (including lip movements and gaze), hand and arm gesture, bodily posture and (to a limited extent) touch. Moreover, extending their use of gesture, humans exchange information through creating and manipulating objects, including objects which themselves represent information, such as images or written text. The natural interactivity vision comes with two claims: (1) that the ease and intuitiveness of human-human-system interaction could be improved tremendously through natural interaction, and (2) that pursuing natural interactivity will lead to completely new families of applications which cannot be realised through traditional graphical user interfaces with a screen, mouse and keyboard.

This talk will present the NITE consortium's analysis of present user needs for a natural interactivity coding tool, and then proceed to present and illustrate the state of ongoing work on the tool's user interface for standard users. The work is based on the following high-level interface specification. The tool should:

• support working with annotation projects, including meta-data handling;
• enable flexible control of raw data files (audio and/or video);
• support annotation of natural interactive communication at any analytical level and across levels/modalities through an existing or new coding scheme;
• enable users to specify new coding schemes, either by editing existing coding schemes or by adding one from scratch;
• enable information extraction and analysis of annotated data.