http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
이두원,최병수 대한전자공학회 2024 전자공학회논문지 Vol.61 No.3
양자 컴퓨팅은 양자역학적 특성을 이용한 신개념 컴퓨팅을 의미한다. 양자 컴퓨팅과 관련한 연구개발은 크게 활용기술 개발과 구현 기술 개발로 구분되며, 다양한 활용기술과 구현 기술이 제안되고 있다. 현재까지 다양한 구현 기술이 제안되고 있으나, 실제로 비트 컴퓨팅의 한계를 뛰어넘기 위해서는 다양한 기술적 이슈들이 해결되어야 한다. 이 과정에서 양자 컴퓨팅의 성능에 영향을 끼치는 모든 구성 요소들에 대한 기능적, 성능적 요구 수준을 명확히 분석하는 것이 매우 중요하다. 이에 따라, 양자 컴퓨팅 성능 평가 기술 연구개발도 최근 급격한 관심을 받고 있다. 관련하여 이미 다수의 양자 컴퓨팅 성능 평가 기술이 개발되었으나, 양자 컴퓨팅 시스템의 모든 구성요소 중 일부분을 주로 다루고 있거나 혹은 분석 가능 수준이 제한되는 한계점이 지적되고 있다. 이에 양자 컴퓨팅 성능 평가 기술들을 체계적으로 비교 분석하는 것이 중요하다. 본 연구에서는 이를 달성하기 위하여 최초로 양자 컴퓨팅 풀 스택 구조 관점에서 분석하는 기술을 적용한다. 다음으로 양자 컴퓨팅 성능 평가 기술들을 양자 컴퓨팅 계산 성능 지표 범위(계산 시간 등)와 도출 능력(분석 시간 등) 측면에서 체계적으로 비교하였다. 결론적으로 본 연구를 통해 양자 컴퓨팅 성능 평가 기술은 아직까지는 풀 스택 구조를 모두 분석하지 못하고 있으며, 분석 기술 측면에서 해결해야 할 기술적 이슈가 많음을 확인하였다. 이를 통해 최종적으로 향후 지향해야 하는 연구개발 방향성을 제시한다. Quantum computing involves the use of quantum mechanical properties in computing. Research and development in this field are divided into utilization technology and implementation technology, with numerous proposals for both. Various implementation technologies have been proposed, but to overcome the limitations of bit computing, several technical issues must be resolved. Therefore, it is crucial to analyze the functional and performance requirements of all components that affect the performance of quantum computing clearly. As a result, there has been significant interest in the research of quantum computing performance evaluation technology. Several quantum computing performance evaluation technologies have been developed, but their limitations have been identified in that they primarily address only some components, analyze only few performance-metric of quantum computing, and restrict the level of analysis. Therefore, it is crucial to systematically compare and analyze quantum computing performance evaluation technologies. To resolve this issue, we propose a comparative method which is based on the quantum computing full stack structure as a new methodological perspective. We then use it to compare systematically various quantum computing performance evaluation technologies using quantitative indicators for computation time, analysis time, coverage of layers, convenience, etc. This study concludes that performance evaluation technology for quantum computing has not fully analyzed the overall quantum computing full-stack structure yet. We also list that there are various technical issues with respect to analysis technology. The study provides insight into the future direction of research and development that should be pursued.
Distributive plural marker tul and its structural licensing
이두원 경희대학교 언어정보연구소 2013 언어연구 Vol.30 No.1
Korean has a copied plural marker -tul (i.e., distributive plural marker, DPM), which looks like a plural morpheme in its form but does not denote genuine plurality of the element it is attached to since it is simply copied from the preceding plural marker. The DPM can be attached not only to a singular noun but also to a verb, adverb, whole verbal chunk or even complementizer, which inherently triggers distributivity, but resists genuine plurality. The licensor of DPMs is not always necessarily a subject but can be a c-commanding nominal. As for the shifted object licensing the DPM attached to the VP adverb, it carries discourse effect such as specificity or presupposed interpretation of distributivity, which induces it to move to the Spec of v*P. Hence, the shifted object c-commands and licenses the DPM attached to the VP adverb. The argument to which the typical distributive markers such as -mata ‘every’, -ssik ‘each’, or hana hana ‘one-by-one’ as well as the plural marker -tul are attached can be a licensor of the DPM. I put forward the observation that in the bi-clausal structure, the DPM attached to the embedded complementizer -ko as well as the matrix adverb is only licensed by the matrix distributive argument. As for the RTOed nominal, it is located in the matrix clause, which induces it to c-command and license the DPM attached to the complementizer and matrix adverb. Like the shifted distributive object, the RTOed argument to which the plural marker -tul is attached also carries discourse effect such as specificity and presupposed interpretation of distributivity, which induces the RTOed object to move to the Spec of v*P in the matrix clause. This set of facts leads us to suggest the condition that the DPM must be locally c-commanded and licensed by PRO, a trace, null argument (i.e, pro) or syntactic argument, which bears distributivity.
여격동사 ‘주’의 영어 대응동사: 자료 중심의 실혐 연구
이두원 한국현대언어학회 2007 언어연구 Vol.23 No.2
In this paper, among other dative verbs in Korean, the verb cwu(ta) 'give' bears a special significance in the sense that the verb allows the goal to be marked by either the dative or the accusative Case, inducing the goal to bear possessor. There are two types of "affection" by the action of the agent to the goal: that is, benefaction and malefaction. The dative verb cwu of the malefaction type is necessarily decomposed into a lexical verb and a lexical causative morpheme; hence, it is a causative verb. In this study, our primary concern is what is the English counterpart of the Korean dative verb cwu 'give' of the malefaction type. It may be 'make', 'give' or a simple transitive verb, according to whether the goal bears possession as an actor or a patient.
이두원 한국중앙영어영문학회 2010 영어영문학연구 Vol.52 No.4
Kes in HiRC constructions cannot be an expletive such as there or it in English. More specifically, kes in HiRC constructions is a demonstrative anaphoric proform referring to the preceding subject or object within the relative clause. Kes in HiRC constructions bears phi-features such as person or number and semantic contents. Kes in HiRC constructions is an anaphoric pronominal: that is, kes in HiRC constructions bears an anaphoric property, even though it is not c-commanded (Kwon 2010). Kes in HiRC constructions can be substituted by the demonstrative anaphor ku casin ‘the himself’ or ku kes ‘the thing’, according to whether it refers to the preceding [+human] or [-human] singular subject or object within the relative clause. Principle A cannot regulate the demonstrative anaphor ku casin. In this vein, it has been shown that the demonstrative anaphoric proform kes in HiRC constructions is compatible, even though it is not c-commanded by its precedent: that is, the demonstrative anaphoric kes in HiRC constructions is Principle A-free. While the donkey anaphora can appear as a pronominal in English, kes in HiRC constructions cannot be replaced by the third person pronoun. Unlike the noun or pronoun kes, kes in HiRC constructions can refer to the respected or the noble, since it can bear the same syntactic and semantic properties as its honorific precedent.
The English Dative Verbs vs. the Korean Dative Serial Verb Constructions
이두원 한국중앙영어영문학회 2007 영어영문학연구 Vol.49 No.3
The so-called serial verb constructions (SVCs) permit Case-alternation, causing the goal to bear possession. For example, while the verb sa ‘buy’ in Korean is a simple transitive verb, the verb buy in English is a ditransitive verb, causing the goal to bear possession; hence, the genuine counterpart of buy is not sa, but sa-cwu ‘buy-give’. This helps us to cultivate the idea that the Korean counterparts of the English ditransitive verbs are the SVCs such as sa-cwu ‘buy-give’. This paper is based on Lee’s (2007b) argument the Korean counterpart of the English ditransitive verb give is the verb cwu base-generated on v in Chomsky’s (1995, 1999) vP-structure (cf. Harley 2002, Jung and Miyagawa 2004). The contribution of the v head is the change-of-state predicate CAUSE (Harley 2002), which induces the goal to bear possession; hence, the verb cwu is a causative verb. This is in accordance with the real generalization that only causative verbs in Korean allow the accusative/dative alternation (Jung and Miyagawa 2004). The transitive verb on V undergoes verbal suffixation to the causative verb cwu on v; hence, a dative (causative) SVC.
Economic Impact of Korea’s Outward FDIs into Developed and Developing Economies across Industries
이두원,허현승 한국무역학회 2009 Journal of Korea trade Vol.13 No.2
Along with the growth of Korea’s outward foreign direct investments (FDI), the concern for the so-called ‘hollowing-out’ effect has grown as well. This paper aims to empirically assess the impacts of Korea’s FDI on Korea’s exports and employment by using time-series and panel analyses. Particularly, the paper compares the impact of Korea’s FDI into developed countries with the impact of Korea’s FDI into developing countries. The manufacturing sector is categorized into 11 sub-industries to account for the possible industry-specific effects. According to the time series analysis, the FDI in China caused exports in several capital-intensive industries. According to the panel analysis, FDI induced more exports, and this effect is particularly strong among Korea’s FDIs in developing economies. In particular, Korea’s FDI in China has induced more exports, but Korean domestic employment has suffered. Employment problems like this are more severe in labor-intensive industries.
李斗源 충주대학교 1994 한국교통대학교 논문집 Vol.29 No.1
English adjuncts why/how and Japanese ones naze/doo behave differently from other wh-adjuncts in multiple wh-questions. English adjuncts show the ECP effect in Syntax. But neither Korean wh-adjuncts except way nor japanese wh-adjuncts except naze show the CNPC effect or the ECP effect. Korean wh-adjuncts way/ettekhey and japanese ones naze/doo cannot appear as the leftmost wh-expression in multiple wh-questions. The sentence that Korean way appears as the leftmost wh-expression in the multiple wh-question is wholly ungrammatical. In English and Korean the words representing why are VP-external, while the words representing how are VP-internal. It is assumed that the degree of dependency on a verb is represented by the hierarchical order in the phrase structure.
Dative Arguments in Unaccusative and Passive Constructions
이두원 한국외국어대학교 언어연구소 2012 언어와 언어학 Vol.0 No.56
The dative arguments with thematic roles of experiencer and possessor in the so-called dyadic unaccusative predicates, which are regarded as a quasi-transitive predicate with a vP phase here, trigger an intervention effect when the nominative objects move to the subject position above the dative argument, so inducing the unacceptable sentences. When a vP functions as a phase, the VP constituent will undergo transfer to the phonological and semantic components and thereafter cease to be accessible to further syntactic operations. At this point, the nominative object within the VP cannot move to the subject position. However, the dative argument does not trigger the intervention effect in the so-called typical unaccusative construction. Also in the passive construction, the complement can undergo movement to the subject position across the dative argument. At this point, the vP is not a phase because it is unaccusative by virtue of having no thematic external argument (e.g., no agent, no experiencer). When the syntactic derivation proceeds by merging the resulting vP with T, the VP-dative argument or nominative complement can undergo movement to Spec-TP for an EPP reason; hence, no intervention effect.
Milestone for Null Pronominalization in the Ciman-Construction: A Reply to An (2013)
이두원 한국언어학회 2013 언어 Vol.38 No.3
The marker nun attached to the relevant nominal in the subsequent clause of the ciman construction is not a plain topic one, but rather a contrastive topic one (Lee 2013a). This induces its contrastive counterpart to be retrieved out of the preceding clause. If the counterpart is identified and may be syntactically independent of its following nominal, the next step is that the nun-marked remnant allows null pronominalization. Beyond the multiple case marking construction (i.e., MCM) context, this may even be extended to the non-MCM context. The nominative- or accusative-marked genuine possessor or the adjunct, which can be replaced by the genitive-marked element, acts separately from its immediately following nominal in the sentence-level context, which induces the nun-marked remnant in the second clause to allow null pronominalization. This is also what the nun-marked goal in the subsequent clause of the ciman construction allowing the double accusative construction shows. As a result, the null pronominalization of the Minor Argument (i.e., null pronominalization of sonkalak-ul 'finger-ACC' in Tom-uy sonkalak-ul 'Tom-GEN finger-ACC' which is Minor Argument Pronominalization (i.e., MAP) in An's (2012a) term) in the MCM context and the other null pronominalization in the non-MCM context are both the same null argument phenomenon even though the former is involved in the nun-marked remnant's ownership of the Minor Argument.
Negative Imperatives Revisited
이두원 한국언어학회 2014 언어 Vol.39 No.1
At PF, when the merger of the morphological features takes place in Korean, the negative imperatives are ungrammatical, if negation blocks the [imp] feature in C from merging with the verb. Short/long form Neg (hereafter, SFN/LFN) is a head of NegP between vP and TP and mal 'don't' in negative imperatives is a lexical spell-out of a combination of Neg and [Imp-Op] in C. Hence, while mal-negative imperatives are possible, short/long form negative imperatives are not. SFN is a syntactic construction, not a prefix attachment. In this vein, the SFN imperative of the morphological causative construction is ungrammatical. When negative prefixes such as pwul, pi, and mi are attached to a predicate, imperatives can be formed with the predicate because there is no functional head Neg which blocks the [Imp] feature in C from merging with the verb. These suppletive forms such as molu(-ta) 'not know' and eps(-ta) 'not exist' behave like other regular short-form syntactic negations (Chung 2007b). This is why the imperatives of suppletive negations are ungrammatical.