Supporting Communication in Spatially Distributed Groups

Beschreibung

Natural Language Generation (NLG) systems in interactive settings often face a multitude of choices, given that the communicative effect of each utterance they generate depends crucially on the interplay between its situational circumstances, addressee and interaction history. This is particularly true in interactive and situated settings. Traditionally, the generation process has been divided into distinct stages of decision making, e.g. content selection, utterance planning and surface realisation. However, this sequential model does not account for the interdependencies that exist among these stages, which in practice can become manifest in inefficient, ineffective communication and an increased cognitive load for the user.This book presents a joint optimisation framework for NLG in dialogue that is based on Hierarchical Reinforcement Learning and learns the best utterance for a context by optimisation through trial and error. The joint model considers decisions at different NLG stages interdependently and produces more context-sensitive utterances than a model that considers decisions in isolation. To enhance the human-likeness of the presented framework, we integrate graphical models, trained from human data, as generation space models for natural surface realisation. The proposed technique is evaluated in a study with human participants and confirms that the hierarchical learner is able to learn an adaptive policy that leads to smooth and successful interactions. Results also suggest that a joint optimisation leads to substantially higher user satisfaction and task success and is better perceived by human users than its isolated counterpart.
Spontaneous and personal encounters and arising communication have a strong impact on work groups. In this thesis, the author develops methods to apply ambient intelligence and qualitative spatial and temporal reasoning to address the spatial gap in distributed work groups. This also contains the problem of privacy in the context of ambient intelligence as privacy is regarded as one of the key acceptance factors. Thus, the aim of this thesis is twofold: the creation of an ambient intelligence proving the ability to support spontaneous encounters and communication in a spatially distributed work setting and an adequate method to ensure privacy in this kind of system in order to increase its acceptance. This cumulative dissertation consists of 6 peer-reviewed papers and one manuscript under review (at the time of submission of this thesis). The contribution of this thesis is a conceptualization of privacy in the context of ambient intelligence including representation, reasoning, and evaluation methods for systems. In addition, an ambient intelligence laboratory is developed. This laboratory is used to address functionality for enabling spontaneous communication in spatially distributed environments and to provide possibilities to investigate ambient intelligence and related systems.

Autor*in

Jasper van de Ven

Themen in »Supporting Communication in Spatially Distributed Groups«

Ambient Intelligence Artificial Intelligence Communication Privacy QSTR

Stimmen zu »Supporting Communication in Spatially Distributed Groups«

Details

Link teilen