<?xml version="1.0" standalone="yes"?>
<Paper uid="P89-1009">
  <Title>COOKING UP REFERRING EXPRESSIONS</Title>
  <Section position="5" start_page="0" end_page="0" type="metho">
    <SectionTitle>
THE REPRESENTATION OF
INGREDIENTS
</SectionTitle>
    <Paragraph position="0"> In most natural language systems, it is assumed that all the entities in the domain of discourse are singular individuals. In more complex domains, such as recipes, this simplification is of limited value, since a large proportion of the objects we find are masses or sets, such as those described by the noun phrases two ounces of salt and three pounds of carrots respectively.</Paragraph>
    <Paragraph position="1"> In order to permit the representation of entities such as these, EPICURE makes use of a notion of a generalized physical object or physob\]. This permits a consistent representation of entities irrespective of whether they are viewed as individuals, masses or sets, by representing each as a knowledge base entity (KBE) with an appropriate structure at.</Paragraph>
    <Paragraph position="2"> tribute. The knowledge base entity corresponding to three pounds of carrots, for example, is that shown in figure 1.</Paragraph>
    <Paragraph position="3"> A knowledge base entity models a physobj in a particular state. An entity may change during the course of a recipe, as processes are applied to it: in particular, apart from gaining new properties such as being peeled, chopped, etc., an ingredient's structure may change, for example, from set to mass. Each such change of state results in the creation of a new knowledge base entity. Suppose, for example, a grating event is applied to our three pounds of carrots between states so and sl: the entity shown in figure i will then become a mass of grated carrot, represented in state sl by the KBE shown in figure 2.</Paragraph>
  </Section>
  <Section position="6" start_page="0" end_page="68" type="metho">
    <SectionTitle>
BUILDING A REFERRING
EXPRESSION
</SectionTitle>
    <Paragraph position="0"> To construct a referring expression corresponding to a knowledge base entity, we first build a deep se-</Paragraph>
    <Paragraph position="2"/>
    <Paragraph position="4"> mantic structure which specifies the semantic content of the noun phrase to be generated. We call this the recoverable semantic content, since it consists of just that information the hearer should be able to derive from the corresponding utterance, even if that information is not stated explicitly: in particular, elided elements and instances of oae-anaphora are represented in the deep semantic structure by their more semantically complete counterparts, as we will see below.</Paragraph>
    <Paragraph position="5"> From the deep semantic structure, a surface semantic structure is then constructed. Unlike the deep semantic structure, this closely matches the syntactic structure of the resulting noun phrase, and is suitable for passing directly to a PATR-like unification grammar. It is at the level of surface semantic structure that processes such as elision and one-anaphora take place.</Paragraph>
  </Section>
  <Section position="7" start_page="68" end_page="69" type="metho">
    <SectionTitle>
PRONOMINALIZATION
</SectionTitle>
    <Paragraph position="0"> When an entity is to be referred to, we first check to see if pronominalisation is possible. Some previous approaches to the pronominalization deci.</Paragraph>
    <Paragraph position="1"> don have taken into account a large number of contextual factors (see, for example, McDonald (1980:218-220)). The approach taken here is relatively simple. EPICURE makes use of a discourse model which distinguishes two principal components, corresponding to Grosz's (1977) distinction between local focus and global focus. We call that part of the discourse model corresponding to the local focus cache memory: this contains the lexical, syntactic and semantic detail of the current utterance being generated, and the same detail for the previous utterance. Corresponding to global focus, the discourse model consists of a number of hierarchically-arranged focua spaces, mirroring the structure of the recipe being described. These focus spaces record the semantic content, but not the syntactic or lexlcal detail, of the remainder of the preceding discourse. In addition, we make use of a notion of discourse centre: this is intuitively similar to the notion of centering suggested by (\]ross, Joshi and Weinstein (1983), and corresponds to the focus of attention in the discourse. In recipes, we take the centre to be the result of  the previous operation described. Thus, after an utterance like Soak the butterbeaa.s the centre is the entity described by the noun phrase the butterbeans. Subsequent references to the centre can be pronominalized, so that the next instruction in the recipe might then be Drain and dnse tltem.</Paragraph>
    <Paragraph position="2"> Following Grosz, Joshi and Weinstein (1983), references to other entities present in cache memory may also be pronominalized, provided the centre is pronominalized. 1 If the intended referent is the current centre, then this is marked as part of the status information in the deep semantic structure being constructed, and a null value is specified for the structure's descriptive content. In addition, the verb case frame used to construct the utterance specifies whether or not the linguistic realization of the entity filling each case role is obligatory: as we will see below, this allows us to model a common linguistic phenomenon in recipes (recipe contezt empty objects, after Massam and Roberge (1989)).</Paragraph>
    <Paragraph position="3"> For a case role whose surface realization is obligatory, the resulting deep semantic structure is then as follows:</Paragraph>
    <Paragraph position="5"> This will be realized as either a pronoun or an elided NP, generated from a surface semantic structure which is constructed in accordance with the following rules: * If the status includes the features \[centre, +\] and \[oblig, +\], then there should be a corresponding element in the surface semantic structure, with a null value specified for the descriptive content of the noun phrase to be generated; t We do not permit pronominal reference to entities last mentioned before the previous utterance: support for this restriction comes from a study by Hobbs, who, in a sample of one hundred consecutive e~.amples of pronouns from each of three very different texts, found that 98% of antecedents were either in the same or previous sentence (Hobbs 1978:322-323). However, see Dale (1988) for a suggestion as to how the few instances of/onc-dbt~a.e pronominalimtion that do exist might be explained by means of a theory of discourse structure like that suggested by Gross and Sidner (1986).</Paragraph>
    <Paragraph position="6">  and \[oblig,-\], then this participant should be omitted from the surface semantic structure altogether.</Paragraph>
    <Paragraph position="7"> In the former case, this will result in a pronominal reference as in Remove them, where the surface semantic structure corresponding to the pronominal form is as follows:</Paragraph>
    <Paragraph position="9"> However, if the participant is marked as non-obligatory, then reference to the entity is omitted, as in the following: Fry the onions.</Paragraph>
    <Paragraph position="10"> Add the garlic ~b.</Paragraph>
    <Paragraph position="11"> Here, the case frame for add specifies that the indirect object is non-obllgatory; since the entity which fills this case role is also the centre, the complete prepositional phrase to the onions can be elided. Note, however, that the entity corresponding to the onions still figures in the deep semantic structure; thus, it is integrated into the discourse model, and is deemed to be part of the semantic content recoverable by the hearer.</Paragraph>
  </Section>
  <Section position="8" start_page="69" end_page="72" type="metho">
    <SectionTitle>
FULL DEFINITE NOUN PHRASE
REFERENCE
</SectionTitle>
    <Paragraph position="0"> If pronominalization is ruled out, we have to build an appropriate description of the intended referent. In EPICURE, the process of constructing a description is driven by two principles, very like Gricean conversational maxims (Grice 1975). The p~'nciple of adequacy requires that a referring expression should identify the intended referent unambiguously, and provide sufficient information to serve the purpose of the reference; and the principle of e~ciency, pulling in the opposite direction, requires that the referring expression used must not contain more information than is necessary for the task at hand. 2 These principles are implemented in EPICUItE 2Similar considerations are discussed by Appelt (1985).</Paragraph>
    <Paragraph position="2"/>
    <Paragraph position="4"> by means of a notion of discriminatory power. Suppose that we have a set of entities U such that</Paragraph>
    <Paragraph position="6"> and that we wish to distinguish one of these entities, zl, from all the others. Suppose, also, that the domain includes a number of attributes (a I, a~, and so on), and that each attribute has a number of permissible values {v,,t, v,,2, and so on}; and that each entity is described by a set of attribute-value pairs. In order to distinguish z~ from the other entities in U, we need to find some set of attribute-value pairs which are together true of zl, but of no other entity in U. This set of attribute-value pairs constitutes a distinguishing descriptior, of xl with respect to the ,~ontext U. A minimal distinguishing description is then a set of such attribute-value pairs, where the cardinality of that set is such that there are no other sets of attribute-value pairs of lesser cardinality which are sufficient to distinguish the intended referent.</Paragraph>
    <Paragraph position="7"> We find a minimal distinguishing description by observing that different attribute-value pairs differ in the effectiveness with which they distinguish an entity from a set of entities. Suppose U has N elements, where N &gt; I. Then, any attribute-value pair true of the intended referent zl will be true of n entities in this set, where n &gt;_ i. For any attribute-value pair &lt; a, v &gt; that is true of the intended referent, we can compute the discriminatory power (notated here as F) of that attribute-value pair with respect to U as fol-</Paragraph>
    <Paragraph position="9"> F thus has as its range the interval \[0,1\], where a value of 1 for a given attribute-value pair indicates that the attribute-value pair singles out the intended referent from the context, and a value of</Paragraph>
    <Paragraph position="11"/>
    <Paragraph position="13"> 0 indicates that the attribute-value pair is of no assistance in singling out the intended referent.</Paragraph>
    <Paragraph position="14"> Given an intended referent and a set of entities from which the intended referent must be distinguished, this notion is used to determine which set of properties should be used in building a description which is both adequate and efficient. 3 There remains the question of how the constituency of the set U of entities is determined: in the present work, we take the context always to consist of the working set. This is the set of distinguishable entisstrictly speaking, this mechanism is only applicable in the form described here to those properties of an entity which are realizable by what are known as abJolute (or t~tereect/ee or pred~tiee) adjectives (see, for example, Kamp (1975), Keenan and FaRm (1978)). This is acceptable in the current domain, where many of the adjectives used are derived from the verbs used to describe processes applied to entities.</Paragraph>
    <Paragraph position="15"> ties in the domain at any given point in time: the constituency of this set changes as a recipe proceeds, since entities may be created or destroyed. 4 Suppose, for example, we determine that we must identify a given object as being a set of olives which have been pitted (in a context, for example, where there are also olives which have not been pitted}; the corresponding deep semantic structure is then as in figure 3.</Paragraph>
    <Paragraph position="16"> Note that this deep semantic structure can be realized in at least two ways: as either the olives which have been pitted or the pitted olives.</Paragraph>
    <Paragraph position="17"> 4A slightly more sophisticated approach would be to restrict U to exclude those entities which are, in G rosz and Sidner's (1986) terms, only present in closed focus spaces. However, the benefit gained from doing this (if indeed it is a valid thing to do) is minimal in the current context because of the small number of entities we are dealing with.</Paragraph>
    <Paragraph position="19"> Both forms are possible, although they correspond to different surface semantic structures. Thus, the generation algorithm is non-deterministic in this respect (although one might imagine there are other factors which determine which of the two realizations is preferrable in a given context}. The surface semantic structure for the simpler of the two noun phrase structures is as shown in figure 4.</Paragraph>
  </Section>
  <Section position="9" start_page="72" end_page="72" type="metho">
    <SectionTitle>
ONE ANAPHORA
</SectionTitle>
    <Paragraph position="0"> The algorithms employed in EPICURE also permit the generation of onc-anaphora, as in Slice the large green capsicum.</Paragraph>
    <Paragraph position="1"> Now remove the top of the small red one.</Paragraph>
    <Paragraph position="2"> The deep semantic structure corresponding to the noun phrase the small red one is as shown in figure 5.</Paragraph>
    <Paragraph position="3"> The mechanisms which construct the surface semantic structure determine whether one-anaphora is possible by comparing the deep semantic structure corresponding to the previous utterance with that corresponding to the current utterance, to identify any elements they have in common. The two distinct levels of semantic representation play an important role here: in the deep semantic structure, only the basic semantic category of the de scription has special status (this is similar to Wel&gt;her's (1979) use of restricted quantification), whereas the embedding of the surface semantic structure's dcsc feature closely matches that of the noun phrase to be generated. For one-anaphora to be possible, the two deep semantic structures being compared must have the same value for the feature addressed by the path &lt;sere spec type category&gt;.</Paragraph>
    <Paragraph position="4"> Rules which specify the relative ordering of adjectives in the surface form are then used to build an appropriately nested surface semantic structure which, when unified with the grammar, will result in the required one-anaphoric noun phrase. In the present example, this results in the surface semantic structure in figure 6.</Paragraph>
  </Section>
  <Section position="10" start_page="72" end_page="72" type="metho">
    <SectionTitle>
PSEUDO-PARTITI'VE NPS
</SectionTitle>
    <Paragraph position="0"> Partitive and pseudo-partitive noun phrases, exemplified by half of the carrots and three pounds of carrots respectively, are very common in recipes; EPICURE is capable of generating both. So, for example, the pseudo-partitive noun phrase three pounds of carrots (as represented by the knowledge base entity shown in figure 1) is generated from the deep semantic structure shown in figure 7 via the surface semantic structure shown in figure 8.</Paragraph>
    <Paragraph position="1"> The generation of partitive noun phrases requires slightly different semantic structures, described in greater detail in Dale (1989b).</Paragraph>
  </Section>
  <Section position="11" start_page="72" end_page="73" type="metho">
    <SectionTitle>
THE UNIFICATION GRAMMAR
</SectionTitle>
    <Paragraph position="0"> Once the required surface semantic structure has been constructed, this is passed to a unification</Paragraph>
    <Paragraph position="2"> grammar. In EPICURE, the grammar consists of phrase structure rules annotated with path equations which determine the relationships between semantic units and syntactic units: the path equations specify arbitrary constituents (either complex or atomic) of feature structures.</Paragraph>
    <Paragraph position="3"> There is insufficient space here to show the entire NP grammar, but we provide some representative rules in figure 9 (although these rules are expressed here in a PATR-Iike formalism, within EPICURE they are encoded as PROLOG definite clause grammar (DCG) rules (Clocksin and Mellish 1981)).</Paragraph>
    <Paragraph position="4"> Applying these rules to the surface semantic structures described above results in the generation of the appropriate surface linguistic strings.</Paragraph>
  </Section>
class="xml-element"></Paper>