<?xml version="1.0" standalone="yes"?>
<Paper uid="A00-3001">
  <Title>Experimenting with the Interaction between Aggregation and Text Structuring</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> In automatic natural language generation (NLG), various versions of the pipeline architecture specified by Reiter and Dale ((Reiter, 1994) and (Reiter and Dale, 1997)) are usually adopted. They successfully modularise the generation problem, but fail to capture the complex interactions between different modules. Take aggregation as an example. It combines simple representations to form a complex one, which in the mean time leads to a shorter text as a whole. There is no consensus as to where aggregation should happen and how it is related to other generation processes ((Wilkinson, 1995) and (Reape and Mellish, 1999)).</Paragraph>
    <Paragraph position="1"> We think that the effect of aggregation spreads from text planning to sentence realisation. The task of text planning is to select the relevant information to be expressed in the text and organise it into a hierarchical structure which captures certain discourse preferences such as preferences for global coherence (e.g. the use of RST relations (Mann and Thompson, 1987)) and local coherence (e.g.</Paragraph>
    <Paragraph position="2"> center transitions as defined in Centering Theory (Grosz et al., 1995)). Aggregation affects text planning by taking away facts from a sequence featuring preferred center movements for subordination. As a result, the preferred center transitions in the sequence are cut off. For example, comparing the two descriptions of a necklace in Figure 1, 2 is less coherent than 1 because of the shifting from the description of the necklace to that of the designer. To avoid this side effect, aggregation should be considered in text planning, which might produce a different planning sequence.</Paragraph>
    <Paragraph position="3"> Aggregation is also closely related to the task of referring expression generation. A referring expression is used not only for identifying a referent, but also for providing additional information about the referent and expressing the speaker's emotional attitude toward the referent (Appelt, 1985). The syntactic form of a referring expression affects how much additional information can be expressed, but it can only be determined after sentence planning, when the ordering between sentences and sentence components has been decided. This demands that the factors relevant to referring expression generation and aggregation be considered at the same time rather than sequentially to generate referring expressions capable of serving multiple goals.</Paragraph>
    <Paragraph position="4"> In this paper, we are concerned with a specific type of aggregation called embedding, which shifts one clause to become a component within the structure of an NP in another clause. We focus on the interaction between maintaining local coherence and embedding, and describe how to capture this interaction as preferences among related factors. We believe that if these preferences are used properly, we would be able to generate more flexible texts without sacrificing quality. We implemented the preferences I. This necklace is in the Arts and Crafts style. Arts and Crafts style jewels usually have an elaborate design. They tend to have floral motifs. For instance, this necklace has floral motifs. It was designed by Jessie King. King once lived in Scotland.  2. This necklace, which was designed by Jessie King, is in the Arts and Crafts style. Arts and Crafts style jewels usually have an elaborate design. They tend to have floral motifs.  in an experimental generation system based on a Genetic Algorithm to produce museum descriptions, which describe museum objects on display. The result shows that the system can generate a number texts of similar qualities to human written texts.</Paragraph>
    <Paragraph position="5"> 2 Embedding in a GA Text Planner To experiment with the interaction between maintaining local coherence and embedding, we adopt the text planner based on a Genetic Algorithm (GA) as described in (Mellish et al., 1998). The task is, given a set of facts and a set of relations between facts, to produce a legal rhetorical structure tree using all the facts and some relations. A fragment of the possible input is given in Figure 2.</Paragraph>
    <Paragraph position="6"> A genetic algorithm is suitable for such a problem because the number of possible combinations is huge, the search space is not perfectly smooth and unimodal, and the generation task does not require a global optimum to be found. The algorithm of (Mellish et al., 1998) is basically a repeated two step process first sequences of facts are generated by applying GA operators (crossover and mutation) and then the RS trees built from these sequences are evaluated. This provides a mechanism to integrate various planning factors in the evaluation function and search for the best combinations of them.</Paragraph>
    <Paragraph position="7"> To explore the whole space of embedding, we did not perform embedding on structured facts or on adjacent facts in a linear sequence because these might restrict the possibilities and even miss out good candidates. Instead, we defined an operator called embedding mutation.</Paragraph>
    <Paragraph position="8"> It randomly selects two units (say Ui and Uk) mentioning a common entity from a sequence \[U1,U2,...,Ui,...,Uk,...,Uu\] to form a list \[Ui,Uk\] representing an embedding. The list substitutes the original unit Ui to produce a new sequence \[U1,U2,...,\[Ui,Uk\],...,Un\], which is then evaluated and ordered in the population.</Paragraph>
  </Section>
class="xml-element"></Paper>