Abstract
The use of virtual agents in training requires them to have several human-like characteristics; one of these is the ability to appear deceptive. We take work from the psychology literature on cues to deception, with a focus on language-related cues, and examine whether it is possible to use resources from the field of Language Technology to construct scenarios with agents showing cues to deception detectable by human judges, a task that has been shown in a text-only context to be difficult. We show that this detection is in fact possible in the context of virtual agents, and that there are interesting results for individual cues, in particular for dialogue- versus lexical-level cues, and a 'placebo' effect.
| Original language | English |
|---|---|
| Title of host publication | AAMAS 2010 |
| Subtitle of host publication | Proceedings of the 9th International Conference on Autonomous Agents and Multiagent Systems |
| Place of Publication | Richland, SC |
| Publisher | International Foundation for Autonomous Agents and Multiagent Systems |
| Pages | 1539-1540 |
| Number of pages | 2 |
| ISBN (Print) | 9780982657119 |
| Publication status | Published - 2010 |
| Event | International Conference on Autonomous Agents and Systems (9th : 2010) - Toronto, Canada Duration: 10 May 2010 → 14 May 2010 |
Conference
| Conference | International Conference on Autonomous Agents and Systems (9th : 2010) |
|---|---|
| City | Toronto, Canada |
| Period | 10/05/10 → 14/05/10 |