• Proceedings of the Acoustical Society of Korea Conference
• /
• 1994.06c
• /
• pp.351-357
• /
• 1994

At least two projects for English-to-Korean translation have been already in action for the last few years, but so far no attention has been paid to the problem of resolving pronominal reference and a default pronoun translation has been considered instead. In this paper we argue that pronous cannot be handled trivially in an English-to-Korean translation and one cannot bypass the task of resolving anaphoric reference if aiming at good and natural translation. In addition, we propose lexical transfer rules for English-to-Korean anaphor translation and outline an anaphora resolution model for an English-to-Korean MT system in operation.

• Korean Journal of English Language and Linguistics
• /
• v.3 no.1
• /
• pp.89-108
• /
• 2003

As pronouns are resolved with their antecedents, definite NPs may enter into the anaphora-antecedent relations with indefinite NPs. This paper is to provide faster and more efficient computational algorithms by which definite NPs are resolved effectively, For this purpose, this paper extends Chierchia's Binding Theory in Categorial Grammar, and definite NPs are resolved with their antecedents by similar algorithms that are used to reflexive resolution. In these algorithms, the relations between indefinite NP and definite NP are represented with λ-expressions, and definite NPs are resolved with their antecedent by λ-conversions.

• Language and Information
• /
• v.3 no.2
• /
• pp.51-66
• /
• 1999

This paper argues that the de se reading is a composite of the corresponding de re one plus the presupposition that there exists a strong epistemic relation between the object of attitude, linguistically realized as an anaphor, and the attitude-bearer, the referent of the anaphor's antecedent. Close examination of the relation between the attitudes de se and anaphora reveals that the notion de se, as currently understood, should be generalized to explain the choice of anaphors. A taxonomy of anaphors in terms of a general notion like presupposition provides further empirical support for the approach taken here.

• Language and Information
• /
• v.2 no.1
• /
• pp.214-255
• /
• 1998

Close examination of previous constraints on bound anaphora that are designed to directly constrain the distribution of referentially dependent(RD) items shows that no universal structural relation may exist that relates RD items to their antecedents. As an alternative to these constraints, this paper proposes an axiom of semantic interpretation, called the Principle of Referential Autonomy, which dispenses with any pretheoretical notion of grammatical functions or configurational notion like c-command. Together with certain English-specific facts, this principle enables us to infer the ungrammaticality of core examples of strong and weak crossover.

• Koreanishche Zeitschrift fur Deutsche Sprachwissenschaft
• /
• v.1
• /
• pp.189-212
• /
• 1999

In diesem Aufsatz haben wir uns $zun\"{a}chst\;damit\;besch\"{a}ftigt$, unter welchen Bedingungen die Pronomina der dritten Person korrekt verwendet werden. Doch davor haben wir uns darauf einlassen, wann die Pronomina Restriktionen unterliegen. Die Situationen, in denen die betreffenden Pronomina sowohl grammatisch, als auch kommunikativ korrekt gebraucht werden, sind 1) Prasentation des Referenzobjekts durch den Sprecher, 2) $Auff\"{a}lligkeit$ des Referenzobjekts, 3) Gemeinsame Fokussierung der Kommunikationspartner auf Referenzobjekte, 4) Warten auf das Referenzobjekt und 5) Erinnerung an Referenzobjekte. In diesen Situationen werden verschiedene Faktoren, z.B. die Lage des Referenten im aktuellen Sprechsituation, fur den Pronominagebrauch mitgewirkt. Seit der Antike bilden die Pronomina einen Gegestand der linguistischen Untersuchung. Seitdem werden sie nach unterschiedlichen Aspekten betrachtet und beschrieben. Vor allem wird der Weg der Forschung zu den personalen, aber auch demonstrativen Pronomina durch zwei Begriffe, namlich Deixis und Anaphora, $ma{\ss}geblich$ bestimmt. Im Hinblick darauf haben wir die Frage diskutiert, ob das $Ph\"{a}nomen$ der Anaphora unter dem der Deixis bzw. der Textualdeixis subsumiert werden kann oder nicht. Man kann dafur $pl\"{a}dieren,\;da{\ss}$ Anaphora und Deixis kommunikativ verschiedene Phanomene darstellen $k\"{o}nnte$. Wir neigen allerdings dazu, $da\ss$ es eine semantische $Integrationsm\"{o}glichkeit$ der Anaphora in die(Textual)Deixis gibt. Nach dem Durchmustern verschiedener Verwendungsweisen und theoretischer Positionen haben wir uns mit der Semantik der Pronomina $befa{\ss}t$, die in der Pronominaforschung bisher weitgehend $vernachl\"{a}ssigt$ worde ist. Dabei stellte sich heraus, $da\ss$ die Beeutung des betreffenden Pronomen vor allem durch die Angabe der $Kommunikationssituationsabh\"{a}ngigkeit$ beschrieben werden kann. Damit wurde ein Versuch unternommen, die festgestellten verschiedenen Gebrauchsweisen semantisch zu systematisieren und somit auf eine einheitliche Semantik $zuruckzuf\"{u}hren$.

• Journal of KIISE:Software and Applications
• /
• v.37 no.12
• /
• pp.928-935
• /
• 2010

Identifying the referents of omitted elements in a text is an important task to many natural language processing applications such as machine translation, information extraction and so on. These omitted elements are often called zero anaphors or zero pronouns, and are regarded as one of the most common forms of reference. However, since all zero elements do not refer to explicit objects which occur in the same text, recent work on zero anaphora resolution have attempted to identify the anaphoricity of zero pronouns. This paper focuses on intra-sentential anaphoricity determination of subject zero pronouns that frequently occur in Korean. Unlike previous studies on pair-wise comparisons, this study attempts to determine the intra-sentential anaphoricity of zero pronouns by learning directly the structure of clauses in which either non-anaphoric or inter-sentential subject zero pronouns occur. The proposed method outperforms baseline methods, and anaphoricity determination of zero pronouns will play an important role in resolving zero anaphora.

• ETRI Journal
• /
• v.43 no.2
• /
• pp.299-312
• /
• 2021

It is necessary to achieve high performance in the task of zero anaphora resolution (ZAR) for completely understanding the texts in Korean, Japanese, Chinese, and various other languages. Deep-learning-based models are being employed for building ZAR systems, owing to the success of deep learning in the recent years. However, the objective of building a high-quality ZAR system is far from being achieved even using these models. To enhance the current ZAR techniques, we fine-tuned a pretrained bidirectional encoder representations from transformers (BERT). Notably, BERT is a general language representation model that enables systems to utilize deep bidirectional contextual information in a natural language text. It extensively exploits the attention mechanism based upon the sequence-transduction model Transformer. In our model, classification is simultaneously performed for all the words in the input word sequence to decide whether each word can be an antecedent. We seek end-to-end learning by disallowing any use of hand-crafted or dependency-parsing features. Experimental results show that compared with other models, our approach can significantly improve the performance of ZAR.

• Proceedings of the Korean Society for Language and Information Conference
• /
• 1998.02a
• /
• pp.184-194
• /
• 1998

• Annual Conference on Human and Language Technology
• /
• 1997.10a
• /
• pp.347-352
• /
• 1997

대용(anaphora)과 생략(.ellipsis)지을 포함한 지시 해결(reference resolution)은 자연어 처리에 있어서 꼭 해결하고 넘어가야 할 문제이며 지시어가 가리키는 피지시어를 올바르게 찾아내는가의 여부가 자연어처리 시스템의 성패를 가름하는 중요한 요소가 된다해도 과언이 아니다. 본 논문에서는 한국어 대용과 생략을 해결하기 위한 기제로서 센터링 이론이 적합한가를 검토하고 나아가 언어에 따라서 다르게 제안된 Cf 목록의 순서를 한국어의 언어특성에 맞게 결정하는 기법을 제안한다. 센터링 이론에서 각 개별언어 간의 차이가 가장 두드러지게 나타나는 부분은 Cf 목록을 구성하는 요소들의 순서이다. 여기에서는 한국어와 언어적인 특성이 유사한 일본어를 대상으로 제안된 순서를 한국어에도 그대로 적용될 수 있는지 검토하고 문제점을 지적함과 동시에 개선 방안을 아울러 살펴보고자 한다.

• Language and Information
• /
• v.4 no.1
• /
• pp.67-82
• /
• 2000

The purpose of the paper is twofold. First, we revise the well-known Centering Theory of anaphora resolution and propose the controlled Information Packaging Theory (for short, CIPT). Second, we suggest a solution to the resolution of the antecedents of pronouns within the framework of CIPT. For this purpose, we select a dialogue of hotel reservation as a domain-restricted discourse, and discourse, and discuss the characteristics of the distribution of pronouns. We suggest that we need to palce the Slot-Link element on the top of the forward centering list. We claim that we need to establish a constraint on conceptual compatibility. As for the pronouns in the global dialogue, we propose a constraint of discourse command.