Generating Tailored, Comparative Descriptions with Contextually Appropriate Intonation

Abstract

Generating responses that take user preferences into account requires adaptation at all levels of the generation process. This article describes a multi-level approach to presenting user-tailored information in spoken dialogues which brings together for the first time multi-attribute decision models, strategic content planning, surface realization that incorporates prosody prediction, and unit selection synthesis that takes the resulting prosodic structure into account. The system selects the most important options to mention and the attributes that are most relevant to choosing between them, based on the user model. Multiple options are selected when each offers a compelling trade-off. To convey these trade-offs, the system employs a novel presentation strategy which straightforwardly lends itself to the determination of information structure, as well as the contents of referring expressions. During surface realization, the prosodic structure is derived from the information structure using Combinatory Categorial Grammar in a way that allows phrase boundaries to be determined in a flexible, data-driven fashion. This approach to choosing pitch accents and edge tones is shown to yield prosodic structures with significantly higher acceptability than baseline prosody prediction models in an expert evaluation. These prosodic structures are then shown to enable perceptibly more natural synthesis using a unit selection voice that aims to produce the target tunes, in comparison to two baseline synthetic voices. An expert evaluation and f0 analysis confirm the superiority of the generator-driven intonation and its contribution to listeners' ratings.

Research through provocation: a structured prototyping tool using interaction attributes of time, space and information.

“…(Resmini & Rosati, 2009). A subjective approach about how information is selected, processed and delivered is the use of the continua (tailored to generic) (White, Clark, & Moore, 2010) common in conversation models and intonation.…”

Context‐Sensitive Natural Language Generation: From Knowledge‐Driven to Data‐Driven Techniques

“…Another system that aims to adapt to an audience at several levels is White et al's () FLIGHTS system. It presents flight‐related information in a way that is particularly tailored towards individual users in terms of content determination, selection of referring expressions, information structure and realisation units for speech synthesis.…”

Ontology Based Interlingua Translation

“…there is one grammar for both realisation and parsing. It means that derivation and generation have the same structure and that we can develop a grammar by testing its correctness in realization in terms of parsing: as a result, we obtain a speed-up in the process of grammar development [14]. Now we first show how to use hybrid logic to model the ontological path of 8, and second we describe a fragment of a CCG for LIS that is able to generate the target translation.…”