File Information
File: 05-lr/acl_arc_1_sum/cleansed_text/xml_by_section/metho/05/i05-6002_metho.xml
Size: 33,893 bytes
Last Modified: 2025-10-06 14:09:40
<?xml version="1.0" standalone="yes"?> <Paper uid="I05-6002"> <Title>Obtaining Japanese Lexical Units for Semantic Frames from Berkeley FrameNet Using a Bilingual Corpus</Title> <Section position="3" start_page="11" end_page="12" type="metho"> <SectionTitle> 2 Proposed Procedure </SectionTitle> <Paragraph position="0"> We used a bilingual corpus (Utiyama and Isahara, 2003) to examine which semantic frames of BFN contained LUs relevant to the Japanese verb osou.</Paragraph> <Paragraph position="1"> JFN, for example, used a mono-lingual corpus to construct the semantic frames. In cases like this, the construction might be inefficient because they have to construct all semantic frames by themselves. But this affects on the reliability of the frames identified and described. This risk of arbitrary description can be reduced by using a bilingual corpus, if it is of high-quality.</Paragraph> <Section position="1" start_page="11" end_page="11" type="sub_section"> <SectionTitle> 2.1 Identifying English equivalents of &quot;osou&quot; </SectionTitle> <Paragraph position="0"> We chose Japanese-English alignments from the bilingual corpus in which the Japanese text contained osou, i.e., the target verb. We obtained 135 alignments from the corpus.</Paragraph> <Paragraph position="1"> The bilingual corpus is consists of two subcorpra. One subcorpus is made of one-to-one alignments. Another is of one-to-many alignments. In the latter, one Japanese sentence is aligned with several English sentences.</Paragraph> <Paragraph position="2"> In the first case, it was straightforward to specify an English word or phrase that translated the target verb, osou. In the second case, however, it is not. So, we singled out an English sentence that corresponds to a Japanese sentence that contained osou. In this process, the identification of osou's English translations was done manually.</Paragraph> <Paragraph position="3"> After this procedure, the following five verbs were identified as English translations of osou: assault, attack, hit, pound, and strike4.</Paragraph> </Section> <Section position="2" start_page="11" end_page="12" type="sub_section"> <SectionTitle> 2.2 Identifying relevant semantic frames </SectionTitle> <Paragraph position="0"> Based on these five verbs, we extracted semantic frames using FrameSQL (Sato, 2003). Semantic frames with LUs that included any of the five verbs were chosen from the BFN semantic frame database (referred to here as BFN).</Paragraph> <Paragraph position="1"> Corpora Mailing List, just one week before the submission deadline. This means that we had little chance to know about the project unless we were &quot;insiders.&quot; 4There were a few other verbs or constructions that served as English translations of osou in the alignments: for example, besiege, engulf, feel pain, occur, hurt, kill, rob, shoot, stab, suffer, wreak on were used as its translations. But we filtered out those less frequent items (whose frequency is less than 3) for purposes of simplicity.</Paragraph> <Paragraph position="2"> Based on Frame Semantics (Fillmore, 1982), BFN posits that a semantic frame is an organization of &quot;semantic roles,&quot; which BFN terms as &quot;Frame Elements&quot; (FEs). Usually, LUs are instantiations or lexical realizations of FEs. Thus, an LU in a frame, F, is a word, or phrase, that, according to the assumptions of Frame Semantics, &quot;evokes&quot; frame F. The definition of the <Attack> frame in the BFN database is used in Figure 1to illustrate the procedure. As indicated, assault, attack and strike are listed as LUs of the <Attack> frame.</Paragraph> <Paragraph position="3"> After manually examining all the semantic frames thus obtained, the five BFN frames were recognized as relevant to the various senses of the target word osou: 1. <Attack> ; 2. <Cause harm></Paragraph> </Section> </Section> <Section position="4" start_page="12" end_page="18" type="metho"> <SectionTitle> 3. <Experience bodily harm> 4. <Cause impact> 5. <Impact> </SectionTitle> <Paragraph position="0"> Semantic frames in the BFN database are supposedly related to one another. There are various relationships, some of which are sometimes encoded by establishing explicit &quot;frame-to-frame relations&quot; (such as &quot;is used&quot; relation) between two frames. Using this information, we obtained the following relationships between the five frames: 1. <Attack> ; 2. <Cause harm> , is used: <Experience bodily harm> ; 3. <Cause impact> , uses: <Impact></Paragraph> <Section position="1" start_page="12" end_page="12" type="sub_section"> <SectionTitle> 2.3 Identifying relevant frame-evoking LUs </SectionTitle> <Paragraph position="0"> in English Each semantic frame has a number of FEs, each of which has lexical realizations, which called LUs. In the work reported here, only verbal LUs were selected as relevant from the English LUs made available in the BFN database.5 Admittedly, there 5 On this point, we recognize a certain kind of discrepancy between the theory and the practice in the BFN framework. If a LU is, according to its defintion, a lexical realization of a certain FE of a certain frame, more nominals should be identified and listed as LUs. For example, in Jack ordered a hamburger at McDonald's, hamburger is a noun that evokes the <Cooking creation> frame. While the <Selling> frame is evoked by order.v, this means that, according the definition of LU, hamburger.n needs to be identified as an LU of the <Cooking creation> frame; more specifically, it is an LU that instantiates the <Food> FE of the frame. It is obvious that the QUALIA STRUCTURE (Pustejovsky, 1995) of hamburger.n contains information of this sort. We suspect that this aspect of &quot;frame-evocation by nominals&quot; does not seem to be properly recognized and coded, and that BFN's current practice of mostly identifying predicates as LUs is somewhat misleading, if we could say so, because it conare a few nominal LUs in certain frames in the BFN, but we ignored them because they found them to be less relevant to our specific task.</Paragraph> <Paragraph position="1"> After identifying all the relevant LUs for the three frames above, we obtained all the English verbs that translated the senses of the target word osou identified in terms of Frame Semantics.</Paragraph> <Paragraph position="2"> For example, the relevant LUs for the <Attack> frame are the following verbs: ambush, assault, attack, charge, invade, jump, lay, set, storm, and strike As was the case with the <Attack> frame, we extracted the relevant LUs for the <Cause harm> and <Cause impact> frames. We manually merged the extracted LUs, and obtained 93 verbal LUs relevant to the Japanese verb osou.</Paragraph> </Section> <Section position="2" start_page="12" end_page="13" type="sub_section"> <SectionTitle> 2.4 Obtaining LU candidates for Japanese FEs </SectionTitle> <Paragraph position="0"> Noun Freq.</Paragraph> <Paragraph position="1"> jiken (incident) 39 boukou (criminal assault) 32 josei (woman) 28 taiho (arrest) 23 hikoku (accused, defendant) 21 yougi (charge, suspicion) 20 kougeki (attack) 20 shounen (boy) 14 tero (terrorism) 14 shougai (injury) 13 higai (damage, harm) 12 kenkei (prefectural police department) 12 manshon (apartment) 12 butai (military unit) 10 fujo (girl and woman) 10 Using the bilingual corpus again, we gathered alignments that had English texts containing the English LUs specified in the way previously described. We obtained 262 alignments. This procedure defined a set of Japanese sentences containing Japanese words or phrases that were natural translations of the LUs in the BFN. ceals the fact that there can be, and actually are, many kinds of frame-evoking effects. BFN has been concentrating on identifying LUs for &quot;governors,&quot; not LUs for the entire set of FEs, for whatever reason. In this respect, it is crucial to note that not all frame-evokers are frame-governors: hamburger.n clearly evokes the <Cooking creation> frame, but there the noun does not govern the <Cooking creation> frame. Arguably, it is unreasonable and even gratuitous to posit the <Hamburger> frame to make hamburger.n a governor. Attack Definition: An Assailant physically attacks a Victim (which is usually but not always sentient), causing or intending to cause the Victim physical injury. The Weapon used by the Assailant may also be mentioned, in addition to the usual Place, Time, Purpose, and Reason. Sometimes a location is used metonymically to stand for the Assailant or the Victim, and in such cases the Place FE will be annotated on a second FE layer.</Paragraph> <Paragraph position="2"> As soon as he stepped out of the bar he was SET upon by four men in ski-masks. Is he INVADING Iraq just to cover other shortcomings? Then Jon-O's forces AMBUSHED them on the left flank from a line of low hills. FEs: Core: Assailant [Asl] The person (or other self-directed entity) that is attempting physical harm to the Victim. The mysterious fighter ATTACKED the guardsmen with a sabre. Victim [Vic] This FE is the being or entity that is injured by the Assailant's attack. The mysterious fighter ATTACKED the guardsmen with a sabre.</Paragraph> </Section> <Section position="3" start_page="13" end_page="14" type="sub_section"> <SectionTitle> Lexical Units </SectionTitle> <Paragraph position="0"> ambush.n, ambush.v, assail.v, assault.n, assault.v, attack.n, attack.v, charge.n, charge.v, fall.v, incursion.n, invade.v, invasion.n, jump.v, lay ((into)).v, offensive.n, onset.n, onslaught.n, raid.v, set.v, storm. v, strike.n, strike.v Created by infinity on Fri Nov 22 14:05:22 PST 2002 It should be noted, however, that there is no established method of recognizing these units automatically; they are part of a text without being marked as such. To solve this problem, we hypothesized that their statistical properties in the texts could be used to pick them up; i.e., we assumed that these LUs were relatively specific to these types of texts and would appear at higher frequencies than usual in the collected text.</Paragraph> <Paragraph position="1"> We collected nouns with higher frequencies under this assumption using a KH Coder 6.</Paragraph> <Paragraph position="2"> The results were sorted according to the parts of speech. The high-frequency nouns thus obtained are listed in Table 1.</Paragraph> <Paragraph position="3"> This provided little information about the semantic classification of the nouns because there was no indication of the LUs that they instantiated. Semantic groupings are latent, however. This meant that we were able to &quot;cluster&quot; the nouns based on certain generic properties to obtain an initial approximation of these groupings. We used a tool called msort (standing for &quot;meaning sort&quot;) (Murata et al., 2001) to establish generic, domain-independent semantic Nouns occurring more than three times were obtained, as shown below:9 human dansei (man), danshi (boy), josei (woman), fujo (woman), joshi (girl), danji (young boy), joji (young girl), youjo (infant girl), shounen (boy), . . .</Paragraph> <Paragraph position="4"> organization kokka (country), gaikoku (foreign country), kokusai (international), sekai (world), . . .</Paragraph> <Paragraph position="5"> product yakubutsu (drug), manshon (apartment), heya (room), keesu (case), naifu (knife), shoujuu (rifle), . . . validity of the proposed method. This evaluation is clearly based on a misunderstanding: the semantic association, or conceptual dependence, between the <Assailant> and the <Victim> FEs is already encoded when we collected only sentences whose main verbs are osou (in Japanese texts) or its translations (in English texts). What we have done with msort is to get subgroupings given a larger semantic grouping of &quot;harm-causing&quot; at a more generic level. Based on our coding experience, we are sure that subclassfication of a given semantic class is based on &quot;semantic types&quot; rather than semantic roles. To give proper subgroupings of the events that the <Attack> frame is relevant, it is necessary to know whether an <Assailant> is a human ([+human, +animate, . . . ]) or an animal ([[?]human, +animate, . . . ]), or whether a <Victim> is a human ([+human, +animate, . . . ]) or an animal ([[?]human, +animate, . . . ]). If we insist that such subclassifications in terms of semantic types into messy details are irrelevant, we are committing what we meant by &quot;mere generalizations for generalizations,&quot; failing to recognized what is really needed in NLP tasks.</Paragraph> <Paragraph position="6"> body part itai (body), soshiki (organization) plant dansei (man), josei (woman), soshiki (tissue) space genba (field), chiiki (region), mokuteki (purpose), hokubu (northern area), shinai (city center) amount gruupu (group) relation jijou (circumstances), keesu (case), jitai (matter), jiken (incident), ryakushiki (informality), kankei (relationship), mokuteki (purpose), genkou (current), . . . activity jisatsu (suicide), satsugai (slaying), shougai (injury), juushou (serious injuries), ishiki (consciousness), utagai (doubt), yougi (suspicion), sousa (investigation), sousaku (search), shirabe (investigation), . . .</Paragraph> </Section> <Section position="4" start_page="14" end_page="14" type="sub_section"> <SectionTitle> 2.5 Identifying LUs for Japanese FEs </SectionTitle> <Paragraph position="0"> Based on the generic semantic groupings produced by msort, we classified nouns into sub-classes by intution, so that they corresponded to the FEs of the BFN frames in the following way: Recall that a semantic frame is a collection of semantic roles, or FEs. In the case of <Attack> , the frame has two &quot;core&quot; FEs, i.e., <Assailant> and <Victim> , and some other &quot;peripheral&quot; or &quot;noncore&quot; FEs such as <Place> , <Time> , and <Weapon> . Thus, <Attack> denotes a situation in which an agent recognizable as an <Assailant> causes (or tries to cause) some <Harm> or <Injury> to someone or a group of people recognizable as a <Victim> at some <Place> and <Time> , sometimes using an item recognizable as a <Weapon> .</Paragraph> <Paragraph position="1"> This means that all we need to do is to classify the nouns in Table 1 into semantic classes such as <Assailant> , <Victim> , <Place> , <Time> , or <Weapon> , with appropriate subclasses where human assailants are distinguished from nonhuman assailants.10 The groupings provided by msort turned out to be useful for this purpose.11 Using this procedure, the nouns obtained on a frequency-basis for <Attack> were classified into the two core FEs, as follows: 10It is important to note that the target data selection procedure of BFN is biased. For example, they put aside a number of problematic cases like metaphorical expressions, and this is clearly reflected in the current frame definitions. We repeated noticed that metaphorically extended senses of a word were systematically dropped in the current release of BFN. For illustration, the sense of attack.n in heart attack is not described in BFN. Descriptive &quot;gaps&quot; of this sort are clearly undesirable; some specific kinds of mapping problems between English LUs provided in BFN and Japanese LUs arise from this.</Paragraph> <Paragraph position="2"> 11We were sometimes unable to identify an FE for a noun class based solely on the output of msort. In these cases, we looked at its usage in the corpus to determine its FE.</Paragraph> <Paragraph position="4"> robbery/robber), heishi (soldier), hikoku (accused person), butai (military unit), kyoudan (religious group) * <Victim> : danshi (boy), josei (woman), fujo (girl and woman), joshi (girl), danji (young boy), joji (young girl), youjo (infant girl), shounen (boy), shoujo (girl), aite (opponent), nihonjin (Japanese), . . .</Paragraph> </Section> <Section position="5" start_page="14" end_page="15" type="sub_section"> <SectionTitle> 2.6 Advantages of proposed method </SectionTitle> <Paragraph position="0"> Using msort turned out to be more beneficial than anticipated when it came to selecting non-core FEs. msort helped to determine noncore FEs correctly to a certain extent. The <Attack> frame, for example, includes noncore FEs such as <Place> , <Time> , <Purpose> , and <Reason> in addition to its core FEs, <Assailant> and <Victim> .</Paragraph> <Paragraph position="1"> msort automatically groups naifu (knife), raifuru (rifle), and pisutoru (pistol) into the &quot;product&quot; category, which corresponds to the <Weapon> FE.</Paragraph> <Paragraph position="2"> Similarly, it automatically groups chiiki (Regional site), hokubu (northern area), and shinai (Inner city) into the &quot;location&quot; category, which corresponds to <Place> . Thus, part of the FE assignment task can be done automatically using msort.</Paragraph> <Paragraph position="3"> The procedure also produced some interesting results. For example, the proposed method automatically specifies a set of lexical items (or lexical units) that clearly have the frame-evocation effect but that are not properly identified as frame elements of a semantic frame in BFN, either in terms of core FEs or peripheral FEs (= noncore FEs). The semantic groupings that were thus au- null tomatically identified are enumerated below: 1. Names denoting an act(ion) of N (N suru (or sareru)) (&quot;(make) do N&quot;): ranbou (violence), boukou (criminal assault), bouryoku (violence), jikkou (execution), shuugeki (assault), kougeki (attack) 2. Names denoting a state of affairs N (V shita + N) (N that S V ): satsugai (slaying), shougai (injury), goutou (burglary/burglar, robbery/robber), satsujin (murder), sasshou (killing and wounding) 3. Result ((Y ni) V shite, N wo owaseta) (&quot;did V , and inflicted N to Y ): juushou (serious injuries) 4. Parts of the compound words: kyoushuu (assault force) (a part of &quot;assault&quot; force) 5. LUs of crime-related frames resulting from <Attack> : utagai (doubt), yougi (charge, suspicion), sousa (investigation), sousaku (search), shirabe (investigation), kentou (investigation), hanketsu (judgement), . . .</Paragraph> <Paragraph position="4"> A second look at the lexical items in 1 above confirmed that most of these words or phrases can be seen as LUs that realize, in Japanese, some of the FEs of BFN's <Attack> frame.12 As sets of lexical items were not classified automatically, we had to determine all classifications manually.</Paragraph> </Section> <Section position="6" start_page="15" end_page="15" type="sub_section"> <SectionTitle> 2.7 Overall results </SectionTitle> <Paragraph position="0"> When the procedure was applied to <Attack> , <Cause harm> and <Cause impact> , the following Japanese LUs for their major FEs were specified: 1. Core FEs of <Attack> : <Assailant> : dansei (man), goutou (burglary/burglar, robbery/robber), heishi (soldier), hikoku (accused person), . . .</Paragraph> <Paragraph position="1"> <Victim> : danshi (boy), josei (woman), fujo (girls and women), joshi (girl), danji (young boy), . . . 2. Noncore FEs of <Attack> : <Place> : genba (field), chiiki (region), hokubu (northern part), shinai (city center) <Weapon> : naifu (knife), shoujuu (rifle), tanjuu (pistol) 3. Core FEs of <Cause harm> : <Body part> : senaka (back) 4. Core FEs of <Cause impact> : To evaluate our results, we compared them with other Japanese resources and methods for analysis, i.e., IPAL (IPA, 1987) and Nihongo Goi Taikei (a Japanese lexicon) (hereafter called Goi Taikei) (Ikehara et al., 1997), which are widely used lexical resources, and semantic frame analysis by FOCAL (Nakamoto et al., to appear; Kuroda et al., 2004), which is a recent framework being developed with the aim of providing BFN-style semantic annotation and analysis for Japanese independent of the Japanese FrameNet (Ohara et al., 2003).</Paragraph> </Section> <Section position="7" start_page="15" end_page="16" type="sub_section"> <SectionTitle> 3.1 Comparison with Goi Taikei descriptions </SectionTitle> <Paragraph position="0"> Goi Taikei contains detailed information on the predicate-argument structure classified according to usage. Its semantic description of osou is given below: 12For the reason of this argument, see note 5 above.</Paragraph> <Paragraph position="2"> (2) 23 shintai dousa (physical motion) (motion)</Paragraph> <Paragraph position="4"> The word meanings were classified from the properties of osou for nouns related to surface cases of the verb. When we compared the frames in BFN and the description provided by Goi Taikei, and examined how the BFN frames corresponded to the Goi Taikei definitions, we obtained the following relationships: Attack (2) 23 shintai dousa (physical motion) Cause harm (1) 20 zokusei henka (property change) Cause impact (1) 20 zokusei henka (property change) First, we did not obtain the meaning &quot;An unexpected event occurred&quot; like (3) in the Goi Taikei. It was difficult to extract words whose meanings described a manner of action, such as fui wo (by surprise) using this method. It was also insufficient to extract only co-occurring nouns from sentences related to verbs. As might be expected, there was a close relationship between (2) and the <Attack> frame. However, we were unable to find <Assailant> s such as sickness in the BFN FEs. Finally, the <Cause impact> frame and (1) were very similar, except that assailant in (1) includes feelings such as worry or sadness.</Paragraph> <Paragraph position="5"> There was a good correlation between the semantic frame constructed from BFN and the one from Goi Taikei. With this method, however, we met difficulties in extracting frames that did not appear on the surface, such as <manner of action> .</Paragraph> </Section> <Section position="8" start_page="16" end_page="17" type="sub_section"> <SectionTitle> 3.2 Comparison with IPAL descriptions </SectionTitle> <Paragraph position="0"> We compared the frames we obtained with the definitions from the IPA Lexicon (IPA, 1987). Below is an excerpt from the description of osou from IPAL: * Caption: osou001001 Semantic definition: An undesirable thing unexpectedly occurs to someone. Sentence valence pattern: N1 -ga N2 -wo Noun phrase 1: bouto (rioter), goutou (burglary), kuma (bear), sentouki (fighter plane), boufuu (wind storm), jishinn (earthquake), ekibyou (plague), keizai All of the frames obtained from BFN seemed to be classified into the first meaning in IPAL, e.g., there were no BFN frames in which <Assailant> recognized &quot;sickness.&quot; With IPAL definitions, it was difficult to distinguish the difference between The bear attacked the traveler and *An economic crisis attacked the traveler, the latter of which sounds unnatural and quite odd, whereas we can do it with BFN definitions: the former 13A term, &quot;predicate frame,&quot; is used in the IPAL to characterize semantic properties of a predicate. While the idea of predicate frames is somewhat related to semantic frames, predicate frames are not defined as semantic frames in the sense of Frame Semantics/BFN.</Paragraph> <Paragraph position="1"> can be classified as an expression in the <Attack> frame, whereas the latter can not. The reason for this is probably that BFN frames successfully specify the semantic interdependence between the <Assailant> and <Victim> roles, whereas such interdependece is not encoded in the IPAL descriptions. We believe this is one of the strengths of frame-based semantic description.</Paragraph> <Paragraph position="2"> BFN definitions are not detailed enough, however. They face problems when we try to account for the constrast between The shark attacked the swimmer and ?*The shark attacked the bank, for example. The latter sentences doesn't makes sense unless it is reinterpreted some way, while it is straightforward to interpret the first sentence against a predatory situation.</Paragraph> <Paragraph position="3"> In interpreting the second, there is a clear conflict or &quot;competition&quot; between two strong readings: one interpretation (reading 1) is against the situation of <Predation> , where the shark is interpreted as a <Predator> and the bank as a <Prey> . Another (reading 2) is against the situation of <Bank Robbery> , where the shark is interpreted as a <Bank Robber> and the bank as a <Warehouse of Valuables> (or simply as a <Bank> ). If reading 2 wins out, an implicit &quot;type coercion&quot; (Pustejovsky, 1995) takes place to the shark so that the referent of the shark is switched to a human who acts as a <Robber> with a nickname &quot;shark.&quot; If reading 1 wins out, by contrast, another kind of implicit type coercion takes place to the bank so that the referent of the bank is switched to an animal (an instance of fish, dolphin, or whale) which acts as a <Prey> , being called &quot;the bank&quot; for some unclear reasons. The preference of the reinterpretation for reading 2 over the other can be accounted for if we are allowed to say that to find someone being called &quot;shark&quot; is more likely than to find some animal being called &quot;bank.&quot; What this suggests is this: pieces of semantic information that would account for &quot;selectional restrictions&quot; of this sort are not specified in the BFN definitions (yet). Therefore, it can be said that the frames constructed from BFN do not classify all meanings of osou in the same way IPAL does not, but these frames specify some finer-grained, selectional aspects of osou's lexical meaning than the IPAL description. As we will see in the next section, this is one of the strong motivations that a framework called FOCAL has tried to extend the BFN.</Paragraph> </Section> <Section position="9" start_page="17" end_page="18" type="sub_section"> <SectionTitle> 3.3 Comparison with FOCAL descriptions </SectionTitle> <Paragraph position="0"> FOCAL is a theoretical framework for semantic analysis and annotation. Its development has been strongly influenced by BFN, but it also tries to extend BFN's scope of semantic analysis to the next stage.</Paragraph> <Paragraph position="1"> In the case of X-ga Y-wo osou, FOCAL recognizes 15 frames in total, listed in Table 4, specifying their hierarchical organization.14 These frames are identified and classified based on the semantic co-variations between <Harm Cause(r))> X, a special case of <Cause(r)> , and <Harm Experiencer> Y , a special case of <Experiencer> . This is important to note that FOCAL puts more emphasis on the specification of the semantic co-variation between X and Y in terms of semantic features because they are crucial characteristics of a semantic frame, which are not captured in the Goi Taikei and IPAL descriptions, and are not clearly encoded even in the BFN description.</Paragraph> <Paragraph position="2"> In FOCAL, frames are defined as idealized models of situations such as Robbery, Predation, assuming that human understanding is situationbased. The descriptive task of FOCAL, then, is to recognize situations and give adequately detailed descriptions to them. Given R is a set of situation-specific roles {r1, . . . , rn}, which are called semantic roles in BFN. Semantic frames are useful only if they serves as specifications of the co-variations among such Rs.</Paragraph> <Paragraph position="3"> For example, F06, as a subclass of the <Attack> class event is defined as follows:</Paragraph> <Paragraph position="5"> where Zprime = Satisfy (r1(Z), Hunger) There seems to be no English noun that names r1. These are the frames that account for more or less all possible readings of X-ga Y -wo osou. The 14 Space limitation disallowed us to show that the 15 frames thus recognized are nearly optimal to exhaustively specify all the situations against which the senses of osou are determined. This was confirmed by multivariate analyses on psychological experiments (Nakamoto et al., to appear). We regret this because the result would surely have answered the question from one of the anonymous reviewers.</Paragraph> <Paragraph position="6"> (nontemporary, e.g., cancer) G5 F14 harm to Y caused by a disease symptom X (temporary, e.g., heart attack) G5 F15 harm to Y caused by a bad feeling X (temporary, e.g., drowsiness) validity of this claim was confirmed through psychological experiments, and reported in (Kuroda et al., 2004; Nakamoto et al., to appear). The BFN identifies 3 frames relevant to the semantics of osou, while FOCAL uses a total of 15 frames to determine the range of situations against which people understand the sentences whose main verb is osou.</Paragraph> <Paragraph position="7"> The 3 BFN frames have been compared with the 15 frames below to assess how well they correspond to one another: This comparison revealed several differences.</Paragraph> <Paragraph position="8"> First, FOCAL specifies situations that the <Attack> frame applies to in much greater detail, although its descriptions are based on semantic frames like BFN's descriptions are. This is mainly because FOCAL identifies frames in terms of conceivable differences in the &quot;purposes,&quot; or &quot;intended effects&quot; of the <Harm Cause(r)> 15, of which BFN's <Assailant> is a special case. This suggests that BFN frames can be further elaborated according to the subclassification of <Assailant> in terms of its purpose.16 The same is conversely true of <Cause harm> and <Cause impact> frames. These BFN frames need to be generalized so that they include nonhuman, nonintentional agents, which is not done in the current BFN. Better matches would be found if the <Cause harm> and <Cause impact> frames were further classified according to the properties of the <Harm causer> and <Impactor> just as in the <Attack> frame.</Paragraph> <Paragraph position="9"> While FOCAL explicitly groups the F01-F05 frames into G1 and combines it with another group, G2, to yield a more general semantic class {G1, G2}, it is not clear whether BFN captures this hybrid class, since the hierarchical relationships among frames are not sufficiently specified. In fact, the comparison with FOCAL revealed that BFN does not classify the <Assailant> types in as much detail as FOCAL does. According to FO-CAL's assumptions, it is <Assailant> 's <Purpose> (including the &quot;null&quot; value) that defines the differences in otherwise similar situations. To identify such subtle differences is exactly what humans are very good at and computers are not. Specification of information of this kind is one of the serious demands arising from many of the NLP tasks.</Paragraph> <Paragraph position="10"> To conclude, we noted that the granularity of the semantic descriptions provided by BFN, IPAL, Goi Taikei, and FOCAL had the following hierarchy: FOCAL > BFN [?] Goi Taikei > IPAL This suggests that, while BFN is clearly useful for a variety of purposes, its semantic descriptions are not detailed enough, particularly when dealing with the polysemy of relatively frequent words like osou in Japanese or hit in English.</Paragraph> <Paragraph position="11"> While our result is only suggestive at best, let 15This is not the same as BFN's <Harm causer> role, which is much more specific than <Harm Cause(r)> in FO-CAL's sense.</Paragraph> <Paragraph position="12"> 16The question of &quot;where to stop,&quot; addressed by one of the anonymous reviewers, would have been answered if we had enough space to show that those 15 frames/situations are nearly optimal to account for all the semantic classifications reflected in selectional restrictions, as explained in note 14. Clearly, we do not need to identify all semantically possible subclassifications; we just need to identify psychologically real subclassifications.</Paragraph> <Paragraph position="13"> us make a brief comment on some methodological aspects of the BFN framework.</Paragraph> <Paragraph position="14"> Overall, BFN definitions for semantic frames are much more oriented or even &quot;biased&quot; for descriptions of activities intended and caused by human, volitional agents. In fact, BFN took a methodological decision not to include metaphorical uses and other &quot;problematic&quot; uses of words for ease of lexicon-building, thereby sacrificing its descriptive range, causing a problem with biased data coverage, as far as we could see. In the case of osou, for example, there were clearly many examples in which harm is not caused by a human, i.e., cases described by FOCAL frame clusters G2: F06-F07, G3: F08-F11, G4: F12, and G5: F13-F15. Therefore, as far as we are concerned with the viability of the frame-based description of situations that can be expressed using osou in Japanese, the current status of the BFN database is only partially successful in that it successfully captures the class of situations specified by G1.</Paragraph> </Section> </Section> class="xml-element"></Paper>