• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10c
• /
• pp.177-179
• /
• 1998

본 논문은 의미론적 어휘개념에 기반한 한국어 동사 Isa 계층구조 시스템을 이용한 Semantic Network을 구축하며, 이를 활용하여 부사와 동사 간의 공기제약관계 설정에 유효한 개념 분류를 수행한다. 일반적으로 많이 쓰이는 한국어 동사 658개를 대상으로 semantic network을 구축한 결과, SENKOV는 44개의 top node를 가지고 있으며 depth 는 약 2.35이었다. 한국어 동사의 semantic network은 영어에서와 마찬가지로 명사보다 top node의 개수가 많고 depth가 훨씬 더 얕았다. 그리고 성상부사의 selectional restriction에 유효한 개념분류를 하는데 SENKOV를 활용하였다.

• Proceedings of the Korean Society for Language and Information Conference
• /
• 2002.06a
• /
• pp.39-47
• /
• 2002

This paper aims at investigating what semantic interpretation the occurrence of stative and non-statve predicates in the complement clauses of the prepositional attitude verbs make a contribution to, as illustrated in (la-b) and (2a-b). (1) a, John believed that Mary was tired (Ambiguous), b. John believed that Mary walked to school. (2) a. John believes that Mary was sick, b. John believes that Mary is sick.

• Annual Conference on Human and Language Technology
• /
• 1999.10e
• /
• pp.316-323
• /
• 1999

동사의 다의성은 결합되는 어휘에 따른 의미확장으로 설명된다. 본고에서는 한국어 지각동사의 기본의미가 갖는 논항관계를 바탕으로 코퍼스를 이용하여 다른 어휘와의 연여관계를 관찰함으로써 공기하는 어휘를 체계화시키고 기본의미와 의미확장의 실제 사용빈도를 조사하는데 그 의의가 있다.

• Proceedings of the Korean Society for Information Management Conference
• /
• 2001.08a
• /
• pp.29-34
• /
• 2001

영상자료 및 소리자료의 색인과 검색을 위해서는 감정동사 및 감정형용사 등의 감정 어휘를 필요로 한다. 그러나 감정어휘는 그 뉘앙스가 미묘하여 분명한 분류체계가 없이는 체계적인 정리가 불가능하다. 이에 따라 본 연구에서는 국어학과 분류사전의 분류체계를 고찰하고 새로운 감정어휘의 분류방안을 연구하였으며, 감정에 따른 기쁨, 슬픔, 놀람, 공포, 혐오, 분노의 6가지 기본유형을 제시하였다.

• Proceedings of the KSPS conference
• /
• 2003.05a
• /
• pp.97-100
• /
• 2003

The present study was planned to investigate the cortical activation correlated with producing morphologically complex Korean verbs by using. fMRI technique. In this study two derivational affixes and two inflectional affixes were selected: pre-final ending and final ending for inflectional affix and passive affix and causative affix for derivational affix. Two Experiment were conducted. The results of two Experiments suggest a possibility that process of pre-final ending is different from final ending.

• Koreanishche Zeitschrift fur Deutsche Sprachwissenschaft
• /
• v.2
• /
• pp.101-128
• /
• 2000

In der vorliegenden Arbeit wurde der Realisierungsmechanismus der fakultativen $Erg\"{a}nzung$ aufgrund des zweidimensionalen Valenzmodells und des zweistufigen Semantikmodlells behandelt. Dabei wurde die $Fakultativit\"{a}t\;der\;Erg\"{a}nzung$ nicht als ein $Ph\"{a}nomen$ im Lexikon angesehen, sondern als ein $Ph\"{a}nomen$ auf der Ebene der Valenzrealisierung, das als Resultat des Zusammenspiels von valenzunabhangigen Faktoren zu betrachten ist. In der bisherigen Valenzforschung wurde die $Fakultativit\"{a}t\;als\;eine\;ausschlie{\ss}lich$ idiosynkratische Eigenschaft des einzelnen Verbs interpretiert, die man bei jedem Verb erlernen $mu{\ss}$. Gegen diese Auffassung wandte die vorliegende Arbeit ein, dass die $Fakultativit\"{a}t$ keine unvoraussagbare idiosynkratische Eigenschaft des einzelnen Verbs ist: Die $Fakultativit\"{a}t\;der\;Erg\;"{a}nzung$ kann durch $Regularit\"{a}ten$ allgemein vorausgesagt werden. Anhand der statischen Lokalisierungsverben im Deutschen wurde gezeigt, dass die $Fakultativit\"{a}t\;der\;lokalen\;Erg\;"{a}nzung$ von den konzeptuellen Prozessen $abh\"{a}ngt:$ Bei dem $Lokalisierungsproze{\ss}$ ist die Realisierung der lokalen $Erg\"{a}nzung$ notwendig. Dagegen ist die Realisierung der lokalen $Erg\"{a}nzung$ bei dem $Delokalisierungsproze{\ss}$ optional. Wenn man die $Fakultativit\"{a}t$ durch allgemeine $Regularit\"{a}ten\;erkl\"{a}ren$ kann, dann ist es nicht $n\"{o}tig,\;die\;Fakultativit\"{a}t$ bei jedem Verb im Lexikon zu markieren. Dadurch kann das mentale Lexikon enorm entlastet werden. Dies bedeutet $f\"{u}r$ deutschlernende $Ausl\"{a}nder$ eine $gro{\ss}e$ Erleichterung.

• Journal of Information Processing Systems
• /
• v.7 no.3
• /
• pp.459-472
• /
• 2011

The vast number of biomedical literature is an important source of biomedical interaction information discovery. However, it is complicated to obtain interaction information from them because most of them are not easily readable by machine. In this paper, we present a method for extracting biomedical interaction information assuming that the biomedical Named Entities (NEs) are already identified. The proposed method labels all possible pairs of given biomedical NEs as INTERACTION or NO-INTERACTION by using a Maximum Entropy (ME) classifier. The features used for the classifier are obtained by applying various NLP techniques such as POS tagging, base phrase recognition, parsing and predicate-argument recognition. Especially, specific verb predicates (activate, inhibit, diminish and etc.) and their biomedical NE arguments are very useful features for identifying interactive NE pairs. Based on this, we devised a twostep method: 1) an interaction verb extraction step to find biomedically salient verbs, and 2) an argument relation identification step to generate partial predicate-argument structures between extracted interaction verbs and their NE arguments. In the experiments, we analyzed how much each applied NLP technique improves the performance. The proposed method can be completely improved by more than 2% compared to the baseline method. The use of external contextual features, which are obtained from outside of NEs, is crucial for the performance improvement. We also compare the performance of the proposed method against the co-occurrence-based and the rule-based methods. The result demonstrates that the proposed method considerably improves the performance.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.8 no.6
• /
• pp.36-45
• /
• 2009

Recently, there are interested in the automatic traffic flowing and accident detection using various low level information from video in the road. In this paper, the automatic traffic flowing and algorithm, and application of traffic accident detection using traffic management systems are studied. To achieve these purposes, the spatio-temporal relation models using topological and directional relations have been made, then a matching of the proposed models with the directional motion verbs proposed by Levin's verbs of inherently directed motion is applied. Finally, the synonym and antonym are inserted by using WordNet. For the similarity measuring between proposed modeling and trajectory of moving object in the video, the objects are extracted, and then compared with the trajectories of moving objects by the proposed modeling. Because of the different features with each proposed modeling, the rules that have been generated will be applied to the similarity measurement by TSR (Tangent Space Representation). Through this research, we can extend our results to the automatic accident detection of vehicle using CCTV.

• Lingua Humanitatis
• /
• v.2 no.1
• /
• pp.93-127
• /
• 2002

This paper deals with a class of Middle English impersonal constructions that involve verbs of two-place argument structure. As is generally understood, the term 'impersonal' is notoriously murky, and after all those researches that have been performed in this area, quite a few issues still remain controversial. The issues we center around in the present study concern the following two. In the type of impersonal constructions we consider, the two arguments-Cause and Experiencer-are both expressed in oblique case, posing the problem of determining which of them functions as the grammatical subject. The issue, however. is not how an argument in oblique case can be taken as the subject: it is well blown that the so called 'dative subject Experiencer' already occurred in Old English. The real issue is why both of the arguments are syntactically realized as nonnominative. The other issue concerns the 3rd-person singular form of the verb. Here again, the crux of the problem may be blurred by the fact that impersonal construction is often defined as one in which the verb has 3rd-person singular form with no apparent nominative W controlling verb concord. But this definition is more nebulous than clear because the notion 'subjectless' is itself highly controversial. Thus, for an expression like me thinketh that-S, it may well be that the verb thinketh ('seems') is 3rd-person singular because the that-clause is the subject. What should be explained of the data brought up here is why the impersonal verb is 3rd-person singular when neither of the NPs associated with it is 3rd person or singular. I argue that we can account for our paradigm examples by looking upon them as 'mixed construction' in which semantic interpretation conflicts with syntactic parsing as a result of case syncretism and gradual establishment of SVO word order. This amounts to saying that the peculiarities of the construction originate with the confused use of impersonal verbs between the sense of 'give an impression' and that of 'receive and impression.'

• Journal of KIISE:Software and Applications
• /
• v.37 no.12
• /
• pp.946-950
• /
• 2010

Many sentiment categorization systems based on machine learning methods use morphological analyzers in order to extract linguistic features from sentences. However, the morphological analyzers do not generally perform well in a customer review domain because online customer reviews include many spacing errors and spelling errors. These low performances of the underlying systems lead to performance decreases of the sentiment categorization systems. To resolve this problem, we propose a feature extraction method based on simple longest matching of Eojeol (a Korean spacing unit) and phoneme patterns. The two kinds of patterns are automatically constructed from a large amount of POS (part-of-speech) tagged corpus. Eojeol patterns consist of Eojeols including content words such as nouns and verbs. Phoneme patterns consist of leading consonant and vowel pairs of predicate words such as verbs and adjectives because spelling errors seldom occur in leading consonants and vowels. To evaluate the proposed method, we implemented a sentiment categorization system using a SVM (Support Vector Machine) as a machine learner. In the experiment with Korean customer reviews, the sentiment categorization system using the proposed method outperformed that using a morphological analyzer as a feature extractor.