Paper: Probabilistic Labeling for Efficient Referential Grounding based on Collaborative Discourse

ACL ID P14-2003
Title Probabilistic Labeling for Efficient Referential Grounding based on Collaborative Discourse
Venue Annual Meeting of the Association of Computational Linguistics
Session Main Conference
Year 2014
Authors

When humans and artificial agents (e.g. robots) have mismatched perceptions of the shared environment, referential com- munication between them becomes diffi- cult. To mediate perceptual differences, this paper presents a new approach us- ing probabilistic labeling for referential grounding. This approach aims to inte- grate different types of evidence from the collaborative referential discourse into a unified scheme. Its probabilistic labeling procedure can generate multiple ground- ing hypotheses to facilitate follow-up dia- logue. Our empirical results have shown the probabilistic labeling approach sig- nificantly outperforms a previous graph- matching approach for referential ground- ing.