File Information
File: 05-lr/acl_arc_1_sum/cleansed_text/xml_by_section/metho/97/w97-0204_metho.xml
Size: 16,901 bytes
Last Modified: 2025-10-06 14:14:38
<?xml version="1.0" standalone="yes"?> <Paper uid="W97-0204"> <Title>Hanshin Publishing Co., Seoul, South Korea.</Title> <Section position="4" start_page="19" end_page="19" type="metho"> <SectionTitle> 3 Inheritance in Frame Semantics </SectionTitle> <Paragraph position="0"> Of course, speakers of a language know something about the differences and similarities among varions types of commercial transactions, e.g. that buying a small item in a store often involves making change, etc. Strictly speaking, this is &quot;world knowledge&quot; rather than &quot;linguistic knowledge&quot;, but this level of detail is required even to parse sentences correctly, e.g. to recognize the different functions of the PPs in &quot;buy a candy bar with a red wrapper&quot; and &quot;buy a candy bar with a $20 bill&quot; and thus to attach them appropriately.</Paragraph> <Paragraph position="1"> mantics from its parent.</Paragraph> <Paragraph position="2"> More complicated cases require more elaborated frames. Thus, &quot;buy a house with a 30-year mortgage&quot; involves a different frame from buying a candy bar, and entails a slightly different interpretation of the PAYMENT element. The relationship between frames is frequently hierarchical; for example, the frame elements BUYER, SELLER, PAYMENT, and GOODS will be common to all commercial transactions; the purchase of real estate contains all of them and (typically) adds a LOAN and a bank (typically) as LENDER. In Our database, these two frames might be represented as shown in Figure i. s Corpus tagging for a sentence like sentence (2): (2) Susan took out a huge mortgage to buy that new house.</Paragraph> <Paragraph position="3"> would have to recognize Susan as playing slightly different roles in the two associated frames.</Paragraph> <Paragraph position="4"> A similar problem in using labels from frame semantic descriptions in the tagging of corpus lines is due to the fact that separate parts of any single sentence can evoke different semantic frames. Consider the following sentence: (3) George's cousin bought a new Mercedes with her portion of the inheritance.</Paragraph> <Paragraph position="5"> In seeing this sentence merely as an expression evoking the commercial transaction frame, we could begin by tagging the subject of the sentence, &quot;George's cousin&quot;, as the BUYER, and the object, &quot;a new Metcedes&quot; as the GOODS, and the oblique object, &quot;her portion of the inheritance&quot;, marked by the preposition &quot;with&quot;, as the PAYMENT. This could be done in a fairly natural and transparent way, as long as the tags were clearly seen as the names of frame elements specifically related to the head verb &quot;bought&quot; in that sentence. But since the words &quot;cousin&quot; and &quot;inheritance&quot; evoke frames of their own, the same sentence could easily come up in our exploration of the semantics of those words as well. In the case of &quot;inheritance&quot;, for example, the information that it gets used for buying something will make clear that this is an instance of estate-inheritance rather than genetic inheritance (or frame inheritance!), and the phrasing &quot;her portion&quot; fits frame understandings about the distribution of an inheritance among multiple heirs. In other words, if we find ourselves tagging the frame elements of Inheritance in that same sentence, the phrase &quot;George's cousin&quot; would be tagged as an HEIR in that frame.</Paragraph> </Section> <Section position="5" start_page="19" end_page="19" type="metho"> <SectionTitle> 4 Applied frame semantics: a </SectionTitle> <Paragraph position="0"> sample frame description.</Paragraph> <Paragraph position="1"> Tagsets for semantic annotation would be derivable from a database of frame descriptions like the ones in Figure 1 above. We can move to another frame to illustrate how frame-based annotation would be accomplished by considering a few words from the 5We leave out of this account the inheritance of a higher-level EXCHANGE frame in the COMMERCIAL-TRANSACTION fralne, and the means for showing that a completed instance of the REALESTATETRANSACTION scene is a prerequisite to the enactment of the associated</Paragraph> </Section> <Section position="6" start_page="19" end_page="21" type="metho"> <SectionTitle> COMMERCIALTRANSACTION scene. </SectionTitle> <Paragraph position="0"/> <Section position="1" start_page="20" end_page="21" type="sub_section"> <SectionTitle> Health Frame </SectionTitle> <Paragraph position="0"> language of health and sickness and showing how the elements and structure of this frame would be identiffed and described. First, appealing to common, unformalized knowledge of health and the body, the frame semanticist identifies the typical elements in everyday health care situations and scenarios, a process involving the interaction of linguistic intuition and the careful examination of corpus evidence.</Paragraph> <Paragraph position="1"> The first product of this analysis is a preliminary list of frame elements (FEN) from this domain, such as, for instance, those shown in Table 1.</Paragraph> <Paragraph position="2"> We have found it necessary to include all of these elements for our purposes, even though some of them are so closely related that they are unlikely to be given separate instantiation in the same clause. Our justification for distinguishing them is based on the results of corpus research and on comparison of the elements of this frame with those of other related frames. Corpus examples in which WOUND and DISEASE are both instantiated are of course rare, and given this complementary distribution we might be tempted to identify these as variants of a single frame element (which we might call AFFLICTION).</Paragraph> <Paragraph position="3"> But this would prevent us from being able to express certain syntactic and semantic generalizations, such as the fact that while we speak of curing diseases, we do not speak of curing wounds, and we speak of wounds but not diseases as heMing, s eThere might be alternative ways of considering such data. It is conceivable that a description with, say, AFFLICTION as a single role element could be maintained In the specific case of the contrast between WOUND and DISEASE we find in metaphor further support for our decision to keep them separate. Metaphoric uses of &quot;cure&quot; and &quot;heal&quot; tend to take direct objects which are target-domain analogues of DISEASE and WOUND respectively. One of the most common instantiations of the DISEASE complement in metaphorical uses of cure is the word ills, a word which in fact appears to be used only in such metaphorical contexts (in talk about &quot;curing society's ills&quot;, for example); and the direct objects of metaphorical heal tend to be based on the notion of a tear or cut or separation, the words wound and scar first of all, but also such words as r/ft, schism, and breach.</Paragraph> <Paragraph position="4"> For each semantic frame, the process of elucidation involves a series of steps: 1. Identification of the most frequent lexical items which can serve as predicates in this frame, 2. Formulation of a preliminary list of frame elements (encoded we expect as a TEL compliant SGML document using feature structures (Sperberg-McQueen and Burnard, 1994), 3. Annotation of examples from a corpus by tagging the predicate with the name of the frame and its arguments with the names of the FE's designating their roles relative to the predicate (also using SGML markup introduced with software developed for this purpose), 4. Revision of the frame description -- specification of the co-occurrence constraints and possible syntactic realizations in the light of the corpus data, and, 5. Retagging of the corpus examples to fit the re- null vised frames.7 The last two steps will be repeated as needed to refine the frame description.</Paragraph> <Paragraph position="5"> by describing certain distinctions between &quot;cure&quot; and &quot;heal&quot; as involving selectionai restrictions. Our inclination, however, is to maximize the separation of frame elements at the beginning, and to postpone the task of producing a parsimonious and redundancy-free description until after we have completed our analysis. ZIn the context of the FrameNet project, the question of how much text will be tagged is a practical one. Our direct purpose is not to create tagged corpora, but to tag enough corpus lines to allow us to make reliable generalizations on the meanings and on the semantic and syntactic valence of the lexical entries we have set out to describe. Whether we choose to tag more than what we need for our analysis will depend on the extent to which the process becomes automated and the resources available.</Paragraph> <Paragraph position="6"> Identifying the semantic flame associated with a word and the FEs with which it constellates does not, of course, constitute a complete representation of the word's meaning, and our semantic descriptions will not be limited to just this. However, we believe that such an analysis is a prerequisite to a theoretically sound semantic formalization, s While any given frame description could be made more precise for other NLP/AI purposes (such as inferencegeneration), the development of such a formalism is not a central part of our current work.</Paragraph> <Paragraph position="7"> For our present purposes, the adequacy of lists of frame elements such as what we present in Table 1 for the vocabulary domain of health care can be established only if precisely these elements are the ones that are needed for distinguishing the semantic and combinatorial properties of the major lexical items that belong to that domain. An initial formulation of the combinatorial requirements and privileges of a frame's lexical members -- here we concentrate on verbs -- can be presented as a list of the groups of FEs that may be syntactically expressed or perhaps merely implied in the phrases that accompany the word.</Paragraph> <Paragraph position="8"> A Frame Element Group (FEG) is a list of the FEs from a given frame which occur in a phrase or sentence headed by a given word. Table 2 gives examples of such FEGs (including FEGs with only one member) paired with sentences whose constituents instantiate them. For purposes of this discussion, the frame elements are identified here using single letter abbreviations, and the structure of an FEG is shown as being merely a bracketed list. We recognize such a naming scheme is inadequate for a large annotation project, and certainly the representation of FEG structures will have to be more powerful.</Paragraph> <Paragraph position="9"> These, however, are minor problems with technical solutions. We focus below on other major issues we are confronting in interpreting the structure of frames as expressed by FEGs.</Paragraph> <Paragraph position="10"> At the lexicographic level of description we could simply list the full set of FEGs for a given lexical unit. However, in many cases the FEG potential of a verb can be expressed in one or more simplifying formulas, by, for example, recognizing some FEs as optional. Thus, since we find both (H, B} (&quot;The doctor cured my foot&quot;) and {H, B, T} (&quot;The doctor cured my foot with a new treatment&quot;), both sentences are using the verb cure in the same sense, we can represent both patterns in a single formula that treats the T element as an optional adjunct SThere are numerous suggestions, not reviewed here, on how to give full semantic representations (Jackendoff, 1994); (Sowa, 1984); (Schank, 1975), etc.</Paragraph> <Paragraph position="11"> FEG Frame Ele- Example (abbr.) ment Group {H,B,T} HEALER, The doctor treated BODYPART, my knee with heat.</Paragraph> <Paragraph position="12"> TREATMENT (H,D} HEALER, The doctor cured DISORDER my disease.</Paragraph> <Paragraph position="13"> {P} PATIENT The baby recovered.</Paragraph> <Paragraph position="14"> {M,B} MEDICINE, The ointment cured BODYPART my foot.</Paragraph> <Paragraph position="15"> {B} BODYPART HIS foot healed.</Paragraph> <Paragraph position="16"> {W} WOUND The cut rapidly (expressed perhaps as {H, B, (T)}).</Paragraph> <Paragraph position="17"> It will not be quite that automatic, however; further distinctions are needed. For example, while we can agree that the TREATMENT element in the previous examples was merely unmentioned, the omission of the DISEASE element in a sentence like &quot;The doctor cured me&quot; has a somewhat different status: there is clearly some DISEASE that the speaker has in mind, and its omission is licensed by the assumption that its nature is given in the context. That is, a possible &quot;of&quot; phrase was omitted from that sentence because its content had been previously mentioned or could otherwise be assumed to be known to both conversation participants. In the tagging of corpus lines, then, we will also indicate the status of &quot;missing&quot; elements to the extent that we can tell what that is. Such information will be presented in the representation of the FEG associated with the predicate. In contrast to cases where frame elements are &quot;missing&quot; (implied but unmentioned, optional, etc.), some examples require that we explicitly recognize (i.e. encode) multiple frame elements for a single constituent. Thus, the disorder may be identified in the description of the patient (e.g. leper, diabetic); we wish to annotate this constituent as Pd, which will be taken as indicating that the constituent satisfies the P role in the frame, but that it also secondarily instantiates a D role, since these nouns designate people who suffer specific diseases (leprosy, degWhere feasible, because of our interest in sortal features of arguments, we will identify the nature of the missing element f~om the context. A similar issue arises in cases of anaphora; we may or may not resolve the anaphora's referent in the annotations, depending on practical considerations of time and effort involved. diabetes). It is important to recognize these cases, since the lexical semantics of verbs sometimes require that certain frame elements be instantiated or clearly recoverable from the context: corpus research on the verb cure, for example, shows that the DIS-ORDER is regularly instantiated. Without explicit coding of the substructure of the PATIENT the sentence He cured the leper ({H,Pd}) would stand as a counter-example to this generalization.</Paragraph> <Paragraph position="18"> There are cases where different but related senses of a predicate have distinct FEG possibilities. For example, the verb heal has two uses, one of which participates in a Causative/Inchoative valency alternation (Levin, 1993) and one which does not. In the use where it refers to the growth of new tissue over a wound, it can be found in both transitive and intransitive clauses: &quot;The cut healed&quot; ({W}) and &quot;The ointment healed the cut&quot; (the ointment facilitated the natural process of healing -- {M, W}). But there is also a purely transitive use with a meaning very close to that of cure, with {H, D} or {M, D}, as in &quot;The shaman healed my influenza&quot; or &quot;The waters healed my arthritis&quot;, and this use of heal usually implies something extra-medical or supernatural. In this usage, there is no corresponding intransitive &quot;*My influenza/arthritis healed.&quot; The verb sense distinctions we make may sometimes be less detailed than those appearing in most dictionaries, since, as many researchers have noted, dictionary sense distinctions are often overprecise and incorporate pragmatic and world knowledge that do not properly speaking inhere in the word itself. An excellent example of this kind of excessive distinction ~ pointed out in (Ruhl, 1989), p.7: one of the dictionary definitions of break is &quot;to rupture the surface of and permit flowing out or effusing&quot; as in He broke an artery. On the other hand, we would expect to capture by this process all the kinds of alternations that (Levin, 1993) has shown to be linked to semantic distinctions, some of them quite subtle.</Paragraph> <Paragraph position="19"> The final versions of the lexical entries will encompass full semantic/syntactic valence descriptions, where the elements of each FEG associated with a verb sense will be linked to a specification of sortal .features, indicating the &quot;selectional&quot; and syntactic properties of the constituents that can instantiate them.</Paragraph> </Section> </Section> class="xml-element"></Paper>