No abstract

In order to pursue research on generating referring expressions in a situated collaboration task, we set up a data-collection experiment based on the Tangram puzzle. For a pair of participants we recorded every utterance in synchronisation with the current state of the puzzle as well as all operations by the participants. Referring expressions were annotated with their referents in order to build a referring expression corpus in Japanese. We provide preliminary results on the analysis of the corpus from various standpoints, focussing on action-mentioning expressions.

No abstract

This paper presents a probabilistic model both for generation and understanding of referring expressions. This model introduces the concept of parts of objects, modelling the necessity to deal with the characteristics of separate parts of an object in the referring process. This was ignored or implicit in previous literature. Integrating this concept into a probabilistic formulation, the model captures human characteristics of visual perception and some type of pragmatic implicature in referring expressions. Developing this kind of model is critical to deal with more complex domains in the future. As a first step in our research, we validate the model with the TUNA corpus to show that it includes conventional domain modeling as a subset.

No abstract