Discourse planning for information composition and delivery: A reusable platform

Abstract

To work effectively in information-rich environments, knowledge workers must be able to distil the most appropriate information from the deluge of information available to them. This is difficult to do manually. Natural language engineers can support these workers by developing information delivery tools, but because of the wide variety of contexts in which information is acquired and delivered, these tools have tended to be domain-specific, ad hoc solutions that are hard to generalise. This paper discusses Myriad, a platform that generalises the integration of sets of resources to a variety of information delivery contexts. Myriad provides resources from natural language generation for discourse planning as well as a service-based architecture for data access. The nature of Myriad's resources is driven by engineering concerns. It focuses on resources that reason about and generate from coarse-grained units of information, likely to be provided by existing information sources, and it supports the integration of pipe-lined planning and template mechanisms. The platform is illustrated in the context of three information delivery applications and is evaluated with respect to its utility.

Automatically generating citizen-focused brochures for public administration

“…The main content is structured in three consecutive parts shown in the middle of the figure; the preparatory content, whose role is to prepare the reader to understand the main content, is structured as a single unit on the far left; the elaboratory content, whose role is to provide additional information, is structured as two units on the near the right; and the procedural content, whose role is to provide information intended to aid the reader in performing an action, is structured as separate units throughout the structure. The contribution of this rhetorical structuring to the text planning process, discussed in Section 5.1, is important for reasons explained elsewhere [4]. Setting the granularity of the information units to section-or-paragraph-sized units proved to be sufficient for the forms of tailoring addressed in the workshop.…”

“…The general approach to NLG adopted for the prototype application is described in detail elsewhere [4] and its unique features with respect to common NLG practice are reviewed in Section 5.1.…”

“…The Centrelink prototype is built using Myriad, a platform for information retrieval and delivery [4]. As discussed in the previous section, it acquires information on the current reader's context and uses this information to plan and generate a brochure that is coherent in expression and personalised for the reader.…”

“…This class has five sub-classes representing the five basic components required of all Myriad-based context models [4]:…”

The pragmatic web

“…These are well-known technologies in the Natural Language Generation (NLG) community [15], and they are built into the Myriad platform on which the DHS-Myriad system is built. Information on the general nature of the Myriad architecture can be found elsewhere [16].…”