Completion of Japanese Sentences 
by Inferring Function Words from Content Words 
Koji KAKIGAHARA and Teruaki AIZAWA 
ATR Interpreting Telephony Research Laboratories 
Twin 21 Bldg. MID Tower, 2-1-61 Shiromi, 
Higashi-ku, Osaka 540 Japan 
Abstract 
A method of generating a Japanese sentence by 
inferring function words from content words using 
valency pa~terns is presented. 
A procedure for selecting an appropriate function 
word, on the assumption that correct content words have 
been selected for a given phrase lattice, is described. A 
method ol ~ inferring a correct verb when verbs are 
recognized less accurately than nouns by the speech 
recognition system is described. Sentences are produced 
from content words as inputs by using the valency 
patterns ohtained from collected dialogue sentences in a 
restricted 1ask domain. Using the semantic features of 
preceding nouns and valency patterns allow a \[airly 
restricted number of candidate verbs to be inferred. 
This method eliminates possible errors at the 
interface between speech recognition and machine 
translation (component technologies of an Automatic 
Telephone Interpretation system) and selects the most 
appropriate candidate from a lattice of typical phrases 
output by 1he speech recognition system. 
1 Background and problems 
An Automatic Telephone Interpretation system is a 
facility which enables a person speaking in one language 
to communicate readily by telephone with someone 
speaking another. Three constituent technologies are 
necessary for such a system: speech recognition, machine 
translation, and speech synthesis. 
Basic research in each of these technologies to 
develop an interpretation system between Japanese and 
English has already started. For this purpose, however, an 
effective interface between speech recognition and 
machine translation is vital because output from the 
speech recognition module will inevitably contain errors 
which the machine translation module cannot accept. 
This paper proposes a method of generatinq a 
Japanese sentence by inferring function words from 
content words using valency patterns. This technique is 
aimed at the selection of the most appropriate candidate 
from a typical phrase lattice that may result from a speech 
recognition system. 
2 Basic assumptions 
In this study the following restrictions relevant to 
the interface problem are assumed: 
(1) A Japanese sentence usually consists of a certain 
number of noun phrases followed by a verb phrase 
at the end. The basic unit of speech recognition is 
assumed to be a continuously uttered phrase, so 
that any input to the machine translation module is 
a 'phrase lattice', i.e., a set of phrase candidates 
hypothesized by the speech recognition module. 
(2) The range of telephone conversation tasks is 
restricted to inquiries from a researcher to a clerk 
about an international conference concerning the 
main topic of the conference, deadlines for paper 
submission, exhibitions, social events, 
accommodation, payment, cancellation, etc.. 
utterance 
results of 
s e.p~h 
r~ 
phrase 1 
'genkou-no' 
'genkou' 8 ~o'--4 I I'mo' 3 I 
L2,oC_  
'kentou' 5 ~ 
'nentou' 1 \[~wo~ I'mo' 3 I 
phrase 2 
'shimekiri-wa' 
'shimekiri' 7 
'shigeki' 3 
IZ _L_I 
L"_""'_LI 2 
Figure 1 
'iku' 
An example of a phrase lattice. 
'itsu-desuka ' J 
'desuka" 4 
"deshita' 3 
"bekika" 1 
"desuka' 4 
'deshita' 3 
"bekika" 1 
'desuka" 4 
'deshita' 3 
"bekika" 1 
291 
3 Research goal 
3.1 A phrase lattice as the result of speech recognition 
Consider a Japanese sentence consisting of two 
noun phrases and one verb phrase: 
'genkou-no shimekiri-wa itsu-desuka'. 
('When is the deadline for a manuscript?') 
Usually a Japanese phrase begins with a jiritsugo- 
word (J-word for short) such as a noun or verb, and ends 
with a sequence of fuzokugo-words (E-words for short) 
such as postpositional particles or auxiliary verbs. In the 
above notation, boundaries between J-words and F- 
words are explicitly indicated by hyphens, and all F-words 
are italicized. 
Figure 1 shows an example of a phrase lattice for 
this sentence obtained as the result of speech recognition. 
Notice that there are candidates for both J-words 
and F-words together with a recognition score of the 
probability that the word is correct. The problem is to 
select the most appropriate candidate from this phrase 
lattice. 
3.2 Selection-by-generation approach 
Attention is focused on candidates for F-words, 
assuming that J-words have already been correctly 
selected by a suitable method. 
The assumption that J-words have been correctly 
selected is realistic if the task domain is limited enough to 
allow a high recognition rate for J-words and a 
knowledge-base, etc. is available for the limited task 
domain. Techniques related to this procedure are now 
being studied. Of the J-words, the predicate atthe end of 
a sentence is less accurately recognized by the speech 
recognition module than nouns. A method to solve this 
problem will be discussed in the second half of this paper. 
In Figure 1, for instance, it is assumed that a 
sequence of J-words: 
'c)enkou' 'shimekiri' 'itsu' 
(manuscript') ('deadline') ('when') 
has been correctly selected according to the recognition 
scores. Corresponding to these J-words, there are three 
sets of candidates for F-words in the phrase lattice: 
'wo' 'wa' 'desuka' 
'too' ' ga" "deshita ' 
'no' "bekika'. 
Our major concern here isthe subproblem of selecting the 
most appropriate one in each of these sets. 
This sub-problem is characteristic of the Japanese 
language. In fact, as easily seen in the above example, 
frequently used F-words, specifically those indicating 
grammatical cases such as 'ga', "wo', "ni', etc., are too 
short to be recognized correctly. Their recognition scores 
are much lower than those of J-words. But in Japanese it 
is often possible to infer the meaning of a given sentence 
from the sequence of J-words when the task domain is 
narrow. 
2P)2 
Our method of selecting the correct F-words is 
composed of two steps: 1) generate a meaningful 
sentence by inferring suitable F-words for a given 
sequence of J-words, and 2) compare these inferred F- 
words with the candidates in the phrase lattice to select 
those most appropriate. 
This idea of 'selection-by-generation' distinguishes 
this approach from previous ones: Hayes et al. \[1\] for 
English or Niedermair \[2\] for German. In this paper only 
Step 1, which is considered the key step, will be discussed. 
4 Generating a sentence by inferring F-words 
The task domain is restricted to inquiries about an 
international conference, and therefore the dialogue is 
basically a repetition of simple questions and answers. 
This increases the probability of inferring the correct F- 
words for each phrase. 
4.1 Key information for the inference 
The following types of information are used to infer 
F-words. The information is described in a lexicon of J- 
words. 
(1) Semantic features of nouns and valency patterns 
First, two types of semantic features are set up for 
nouns appearing in the restricted task domain. One is a 
general type of semantic feature, independent of the task 
domain, such as abstract, action, concrete, human, 
location, time, number and diversity. The other is a 
specific type of semantic feature dependent on the task 
domain. Table 1 shows examples of such features. 
Using these semantic features valency patterns of 
the basic predicates necessary in the task domain are 
defined. As an example, the predicate 'okuru' ('send' in 
English) is given the following valency patterns: 
N\[con/-tra\]'wo' + V, 
N\[con/-tra\]'wa' + N\[Ioc\]'ni' + V, 
N\[con/-tra\]'wa' + N\[hum\]'ni' + V, 
N\[con/-tra\]'wa" + N\[tim/pro\]'madeni' + V, 
N\[tim/pro\]'madeni" + N\[con/-tra\]'wo' + V, 
N\[hum\]'ni'+ N\[con/-tra\]'wo' + V, 
N\[hum\]'ga'+ N\[con/-tra\]'wo' + V, 
N\[hum\]'ga' + N\[con/-tra\]'wo' + N\[hu m/-pro\]'ni' + V, 
N\[hum/-pro\]'wa' + N\[con/-tra\] 'wo' + N\[h um\]'ni'+ V, 
etc.. 
The first valency pattern in this list, for instance, specifies 
that the predicate V ('okuru') can take one noun phrase 
consisting of a noun with the general semantic feature 
'concrete' / specific semantic feature non-'transport' and 
F-word 'we'. 
In this way the valency patterns summarize the basic 
J-word and F-word relationships, and thus give the most 
essential information for inferring F-words from a given 
sequence of J-words. 
gener___a/ 
abstract 
Table 1 Semantic features 
specific example 
logic 'riron'(theory), 'houhou'(method) 
state 'yousu'(state), 'baai'(case) 
'language 'nihongo'(Japanese), 'eigo'(English) 
learning 'bunya'(field), 'senmon'(specialty) 
intention 'kyoumi'(interest), 'kibou'(hope) 
value 'hitsuyou'(necessity) 
sign 'namae'(name) 
labor 'youken'(business) 
concrete document 'genkou'(manuscript),'youshr(form) 
transport 'basu'(bus), 'tikatetsu'(subway) 
article 'syashin'(photograph) 
.___money. 'okane'(money),'kado'(cash card) 
pronoun 'kore'(this) ,'dore'(which) 
human human 'happyousya'(presenter) 
pronoun 'clare'(who) 
location ~ ~kaijyou'(hall),'hoteru'(hotel) 
region ~ 'Kyoto'(Kyoto) ,'kaigai'(foreign) 
po_. sition 'temoto'(hand) ,'aida'(between) 
pronoun 'doko'(where) 
time thee 'jikan'(time), 'ima',(now) 'ato'(after) 
pronoun I 'itsu',(when)'nanji'(what time) 
number amount I 'ninzuu'(the number of people) 
unlt I 'en'(yen), 'doru'(dollar) 
cost \] 'tourokuryou'(registration fee) 
price I 'muryou'(free), 'ikura'(how much) 
act 'sanka'(par ticipation), 'yotei'(plan) 
diverse 'nani'(what) ,'hoka'(else) 
These valency patterns are obtained from the 
valency patterns of predicates (obtained from dialogues 
collected for the task of inquiries about an international 
conference). If necessary, certain modifications such as 
omission of the nominative case, modification of the 
word sequence or of the F-words, and addition of 
interrogative pronouns are carried out. In dialogue 
sentences, the nominative case such as 'watashi' (T) or 
°anata' ('you') is seldom used; hence, the nominative case 
is usually not included in the valency patterns. To describe 
the modification of the two valency patterns for 'okurU' 
('send'): 
N\[tingpro\]'madeni" + N\[con/-tra\]'wo' + V (a) 
N\[con/otra\]'wa' + N\[tim/pro\]'madeni' + V (b) 
If the noun N\[con/-tra\] of the valency pattern (a) becomes 
the subjec% the F-word is replaced with 'wa'and the word 
sequence is changed, often resulting in the valency 
pattern (b). Interrogative pronouns are added to produce 
valency patterns specific to interrogative sentences 
because a large number of questions occur in this task 
domain. 
In a \]imited task domain, even individually optional 
cases behave in a similar way to the obligatory case for 
each predicate. Therefore, the optional cases are 
described in these valency patterns. When valency 
patterns were prepared for 65 words working as 
predicates, an average of about 11 valency patterns were 
produced for each predicate. Details will discussed in 
Chapter 5. 
(2) Connection of two nouns by F-word 'no" 
it is inferred that nouns which cannot be processed 
through valency patterns are likely to be connected with 
the F-word 'no' (roughly corresponding to 'of' in English) 
in the form 'A no B', where A and B denote nouns. For a 
given noun A, the other noun B can also be specified 
through the semantic features. For instance, the noun 
'kaigr ('conference') can be joined with other nouns as 
follows: 
'kaigi' + 'no' + N\[hum/hum\], 
'kaigi' + "no' + N\[abs\], 
'kaigi' + 'no' 4. N\[tim/tirn\], 
'kaigi' + 'no' + N\[Iodins, pos\], 
N\[abs\] + 'no' + 'kaigi', 
N\[num/amo\] + "no' + 'kaigi', etc.. 
As shown above, whether or not to insert the F- 
word "no" is automatically determined by presetting 
which nouns are to be connected with the F-word "no'. 
(3) Syntactic information 
Pure syntactic knowledge is also useful in this 
process. It is known that, in Japanese, no F-word can be 
attached to an adverb or a conjunction, and that a verb in 
conditional form can be connected with an adjective via 
F-words such as'ba'. 
In addition, the following rules are used, for 
example: 
Continuative verb form + F-word + verb 
--> conjunctive particle 'te', 
Conclusive verb form + F-word + verb 
-~ conjunctive particle 'to', 
Continuative verb form + F-word + adjective 
-~ conjunctive particle 'te', 
conjunctive particle " temo ', 
Attributive verb form + F-word + noun 
--7 no F-word, etc.. 
4.2 Outline of the process of inferring F-words 
Figure 2 (a) illustrates a process of inferring F-words 
in a Japanese sentence: 
'kaijyou-de kaigi-no yousu-wo 
rokuonshi-temo ii,desuka', 
(May I record the speech of the conference 
atthe hall?). 
In this case it is assumed that J-words 'kaigi', 'yousu', 
'rokuonshi', and 'ii' are correctly recognizable. 
The inference proceeds as follows: 1) syntactic 
information can connect 'rokuonshi' and 'ii' with F-words 
'te" or "temo' and 'desuka' to generate the phrases 
'rokuonshi-te' or 'rokuonshi-temo' and "ii.desuka', 
'293 
respectively, 2) considering the semantic features of the 
first three J-words, and taking the fourth J-word 
'rokuonsuru' as V, the valency pattern: 
N\[Ioc\]'de' + N\[act,abs\]'wo' + V, 
can be applied to them, 3) there are two possible 
connections: 'kaigi-no yousu' and 'kaijyou-no kaigi', and 
4) considering both 2) and 3) together, sentences: 
'kaijyou-(de, no) kaigi-no yousu-wo 
rokuonshi-(te, temo) ii-desuka', 
can finally be derived. 
In a similar way, (b) shows how the following 
sentence is to be processed: 
'genkou-wa itsu-madeni okure.ba 
yoroshii-desuka ', 
(By what time may I send the manuscript ?). 
Here, 1) 'okure' is combined with 'yoroshii' by the F-words 
'ha' and 'desuka', to yield 'okure-ba yoroshii-desuka', 2) 
analyzing the semantic features of the nouns 'genkou' 
and 'itsu' and the presence of the verb 'okuru', the 
following valency pattern is applied: 
N\[conl-tra\]'ha" + N\[timlpro\]'madeni' + V. 
3) Since the nouns 'genkou' and 'itsu' cannot be 
connected by the F-word "no', 4) the following sentence: 
'genkou-wa itsu-madeni okure-ba 
yoroshii-desuka', 
is finally obtained. 
5 An experiment to produce sentences from valency 
patterns 
Using the valency patterns obtained from collected 
dialogue sentences, we carried out an experiment of 
producing sentences. Of the total of 256 interrogative 
sentences, 146 were used in determining valency 
patterns. The number of verbs was 65 and that of nouns 
was 229. In total, 669 valency patterns were prepared 
(10.7 patterns for each verb on the average). 
In addition to the collected dialogue sentences, we 
prepared 70 test questions. For these interrogative 
sentences, we carried out a sentence-producing 
experiment. The results of this experiment are shown in 
(a) Oriqinal u.tterance 
I 'kaijyou-de kaigi-no yousu-wo rokuonshi-temo ii-desuka' 1 ~ F-words 
t J ~ to be inferred 
Sequence of J-words correctly recoqnized 
noun '/ noun 'l noun , verb adjective 
loclins ' act ' abslsta ', negative or conclusive ~.. i i 
....... _'~__: ..... _~_: L__~___.: ~ c°=t~2u=J~E_, etc. 
2)Valency pattern -~-~N \[i oi~de'-+ N\[act~abs\]';o'TV - - - I--~/- ~ I ~F .... =---\[ ............... 1)Syntactic information 
3,Connection of nouns I '"aij'°u "° "a'=' \[ ............... l 'kaigi n°'°u'u' \]r / \[_.. 'r°V=°"shi'(te' 'e;"°' ii'desli'a' I~,/// 
4)Sentence qeneration j(/ 
/ 'kaiiyou-Cde..o~ k~igi-.o you.u-wo rokuo,shi-Ire, temo~ ii-,esu~a' \[ / 
............... 1 Lexicon of J-words 
part of speech 
conjugation 
semantic features 
(b) Or_~nal utterance 
'genkou-wa itsu-madeni okure-ba yoroshii.desuka'-- I ~F-words 
~to be inferred Sequence of J-words correctly recoqnized 
L'o e°-u' .,,so. iiill-- ,u,e' iiilI  ,- .............. ......... 
, r- ................... r-- ~ ..... ~ ', Lexicon of J-word=s noun ', noun ~ verb adjective \] part of speech 
,m~ con/doc '1 tim/pro. '~ conditional conclusive -~..-~ conjugation \ I _~_ = _~. I semantic features 
...... ~ ........... T--_ ....... f ..... ~ ...... -~-'.-. ...... 2)Valency pattern 
- LN\[c_°_n/d°c\]:wa_'_+N\[ti.m_\]:ma_den!_+V__ i l 
3)Connection.of..n. oun.s. ...................... ; ~re- ba yoroshii.desuka'l)Syntactic informat!onj 
'genkou no itsu' / 
,,Sente=e ........... .... / 
\[ '..n.ou-wa ...-ma,e.. o,.=:~a .oro.,,-,.uka' I 
Figure 2 Process of inferring F-words. 
29L 
Table 2. In this experiment, the sequence of J-words of 
each test sentence was input, and a complete sentence 
including F-words was output. The correct answer rate in 
Table 2 is the percentage of all the output sentences that 
were consi~.~tent with the input sentences. At the first trial, 
64.3% coirect sentences were produced from the 
prepared valency patterns. 
Table 2 Sen1 ences obtained using valency patterns 
Number -- 
of 4 
candidate 5 
sentence:; --~- 
Correct sentences (total) 
Correct answer rate (%) 
0 
I 
Number ---~ 
of .- 
candidate _3 
sentences 4 
5 
6 
Incorrect s~.mtences (tdtal) 
Trial I 2 3 4 i 5 6 ........ 
i 1 12 14 15 16 I 16 16 
__L_ 1___~4 2___2_0 2~_ z___~d__24 24 
3 0 0___0___ s, s --%-_ 
3 3 3 3 3 3 
7 7 7 7 7 7 
6 6 6 6 7 5 
7 0 0 0 0 0 4 
8 2 2 2 2 2 3 
9 0 0 0 0 0 0 
10 1 1 1 1 1 1 
45 53 59 66 67 70 
64.3 75.7 I 84.3 94.3 95.7 100 
81 0 01 0 0 01 
61 6 01 0 0 01 
61 6 71 0 0 01 
11 I 01 0 0 01 
01 0 01 0 0 01 
II I '' I 0 Ol 
31 3 31 3 3 ol 
251 17 111 4 3 Ol 
669 1 677 6831 690 691 69'41 Valency patterns (total) 
The !lpper hatf of Table 2 shows the number of 
candidates in the trials where some of the output 
sentences were correct. For example, the number 7 shown 
for the first trial in the line corresponding to 5 candidate 
sentences means the number of candidate sentences was 
5~ and that 7 of the 70 test sentences were correct ones. 
The lower half of Table 2 shows the number of test 
sentences in'trials where no candidate sentence was 
produced or where none of the candidate sentences 
produced were correct. 
The figures for the second and subsequent trials in 
Table 2 show the change' in the correct answer rate when 
additional valency patterns were used to increase the 
incidence of correct sentences. In this experiment, enough 
valency patterns were added so th'at the sixth trial always 
produced correct sentences. 
AI~ elf the test sentences used in the above 
experiment were simple interrogative sentences. As 
shown in Table 2, tile Use of ~alency patterns allows easy 
production of a complete sentence from a given sequence 
of J-words. 
6 Inferring an omitted verb 
The verb in a given sequence of J-words has an 
important role in this method because it allows the 
selection of a correct valency pattern. It would he difficult 
to proceed by this method if the verb is omitted for some 
reason, such as speech recognition failure, or if it were 
originally omitted as is often the case in Japanese 
dialogue. 
However, in this restricted domain, nouns with 
particular semantic features are often related to 
particular verbs. For example, as shown in Figure 3, in 
sentences which contain a noun with the semantic 
feature of concrete/document, the noun + F-word 'wo' 
tends to be accompanied by the verb 'okuru'('send'), 
'kaku'('write') or 'motsu'('have'), and the noun + F-word 
'hi' tends to be accompanied by the verb 
'kinyuusuru'('enter'). 
This suggests the possibility of inferring an omitted 
verb from the nouns by inversely applying a suitable 
valency pattern. In fact, the definition of a valency 
pattern can be generalized as follows: 
N\[sem Iglsem Is\] + N\[sem2glsem2s\] -t .... + V\[v-class\], 
where V\[v-class\] denotes a verb belonging to verb class 'v- 
class'. This valency pattern can be used to infer a verb V\[v- 
class\] associated with a set of nouns N\[semlglsemls\], 
N\[sem2glsem2s\], etc.. 
This is schematically illustrated in Figure 4. When 
there is a verb group A (consisting of verbs which are 
inferred when a certain F-word is added to noun A) and a 
verb group B (consisting of verbs which are inferred when 
a certain F-word is added to noun B),the common area of 
these two groups indicates the verbs which are inferred 
from the valency pattern containing noun A and noun B. 
For example, in a sentence which contains a noun 
'setsumeisyo' ~erb 
'genkou' ~ ~ ~ ) 
\[concreteldocument "l ) 
'tourokuyoushi' ~ ~/erb ~ ....... \ 
' '~shi' 
Figure 3 Frequent combinationsof r~ouns and verbs 
295 
'moushikomiyoushi' 
'temoto' 
\[somontic  eoturoA J /  oncrete,do  meot 
'WO' 
'ni' 
'aru' 
a' 
Figure 4 Verbs defined by multiple nouns 
with the semantic features of concrete/document and a 
noun with the semantic features of location/position, 
verbs such as 'aru'('be') and 'motsu'('have') tend to be 
selected, and F-words specific to these verbs are chosen. 
Table 3 shows the number of verbs which are 
inferred from a given sequence of nouns using the 
valency patterns described in Chapter 5. Since valency 
patterns were prepared for 65 verbs, the verbs are 
inferred -from these 65. 
The columns in Table 3 Show the number of verbs 
inferred. The lines in the same table show the number of 
nouns in the valency patterns. For example, when the 
number of inferred verbs is 5, there are 6 valency patterns 
where 5 verbs are inferred; and of these 6 patterns 1 has 
one noun, 4 have 2 nouns, 1 has 3 nouns and none have 4 
nouns. 
In counting the number of inferred verbs, only verbs 
having a valency pattern consistent with given valency 
patterns are counted. When the noun of a valency 
pattern bears a specification as to the upper-level general 
semantic features but no specification as to the lower- 
task~dependent semantic features, the verbs of the 
valency patterns bearing a specification as tothe lower- 
evel semantic features are counted. Conversely, for 
valency patterns where the lower-level semantic feature 
is specified, the verbs bearing no specification as to the 
Iower.-levei semantic features are not counted. For 
example, in the followingtwo valency patterns, the #erbs 
Vl and V2 are inferred from the pattern (a), while only 
the verb V2 is inferred from the pattern (b). 
N\[con\] + N\[Ioc\] + Vl (a) 
N\[con/doc\] + N\[Ioc\] + V2 (b) 
As can be seen in Table 3, only one verb was inferred 
in more than 50% of the valency patterns. Irl 90% of the 
remaining valency patterns where multiple verbs were 
in~erred, the number c~f verbs i~fe~rred was 6 o~ le~s. ~hese 
Table 3 Verbs inferred using valency patterns 
1 2 3 4 Total 
1 6 77 29 5 117 
2 7 17 3 0 27 
3 2 13 0 16 
4 3 4 0 0 7 
5 1 4 1 0 6 
6 0 7 0 0 7 
7 1 4 0 0 5 
,..i 
8 o 21 o o 2 
-- i 
9 0 0 0 0 0 
10 2 1 0 0 3 
= 
11 2 0 0 0 2 
12 2 1 0 0 3 
16 1 0 0 0 1 I 
18 1 0 0' 0 1 
21 1 0 0 0 1 
29 1 0 0 0 1 
Total 30 130 34 5 199 
A: Number of verbs inferred 
B : Number of nouns in valency patterns 
results indicate that, in a restricted task domain, the 
semantic features of the preceding nouns and valency 
patterns allow a fairly restricted number of candidate 
verbs to be inferred. 
7 Conclusions 
As a first step toward a better interface between 
speech recognition and machine translation, a method 
which is particularly useful for Japanese sentences was 
proposed to infer F-words for a given sequence of J- 
words. 
In a restricted task domaia., the most appropriate F- 
word can be inferred from a given sequence of J-words if 
the task-dependent semantic features of nouns are preset 
and the information of valency patterns is utilized. 
In addition, the results of this study suggest that 
correct verbs can be inferred from valency patterns. 
The authors are now evaluating the effectiveness of 
the procedures proposed in this paper by applying them 
to actual results of speech recognition. 
Acknowledgment 
The authors are deeply grateful to/Dr. K~rematsu, the president of 
/ATR Interpreting Telephony Research Eaboratories, and all the 
members of ATR Interpreting Te/lephor~y Research Laboratories for 
their constant help and encouragement: 
References 

\[1\] Hayes, P.J.et al., "Parsing Spoken Language; A Semantic 
Caseframe Approach", EOLING,86 

\[2\] Niedermair, G., "Divided a~d Valency-Oriented Parsing in 
Speech Understandir)g", COLING 86. 
