Givenness Hierarchy Theoretic Cognitive Status Filtering
- Poulomi Pal, Computer Science, Colorado School of Mines, Golden, Colorado, United States
- Akshay Swaminathan, Computer Science, Colorado School of Mines, Golden, Colorado, United States
- Lixiao Zhu, Computer Science, Colorado School of Mines, Golden, Colorado, United States
- Andrea Golden-Lasher, Computer Science, Colorado School of Mines, Golden, Colorado, United States
- Tom Williams, MIRRORLab, Colorado School of Mines, Golden, Colorado, United States
AbstractFor language-capable interactive robots to be effectively introduced into human society, they must be able to naturally and efficiently communicate about the objects, locations, and people found in human environments. An important aspect of natural language communication is the use of pronouns. According to the linguistic theory of the Givenness Hierarchy (GH), humans use pronouns due to (subconscious) assumptions about the cognitive statuses their referents have in the minds of their conversational partners. In previous work, Williams et al. presented the first computational implementation of the full GH for the purpose of robotic language understanding, leveraging a set of rules informed by the GH literature. However, that approach was designed specifically for language understanding, oriented around GH-inspired memory structures used to assess what entities are candidate referents given a particular cognitive status. In contrast, language generation requires a model in which cognitive status can be assessed for a given entity. In this work, we present and compare two such models of cognitive status: a rule-based Finite State Machine model directly informed by the GH literature and a Cognitive Status Filter designed to more flexibly handle uncertainty. The models are demonstrated and evaluated using a silver-standard English subset of the OFAI Multimodal Task Description Corpus.