• Journal of KIISE:Software and Applications
• /
• v.30 no.5_6
• /
• pp.570-578
• /
• 2003

Maximum entropy models are promising candidates for natural language modeling. However, there are two major hurdles in applying maximum entropy models to real-life language problems, such as prepositional phrase attachment: feature selection and high computational complexity. In this paper, we propose a maximum entropy boosting model to overcome these limitations and the problem of imbalanced data in natural language resources, and apply it to prepositional phrase (PP) attachment and part-of-speech (POS) tagging. According to the experimental results on Wall Street Journal corpus, the model shows 84.3% of accuracy for PP attachment and 96.78% of accuracy for POS tagging that are close to the state-of-the-art performance of these tasks only with small efforts of modeling.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.10d
• /
• pp.670-672
• /
• 2002

Park과 Zhang은 최대 엔트로피 모델(maximum entropy model)을 실제 자연언어 처리에 적용함에 있어서 나타날 수 있는 여러가지 문제를 해결하기 위한 최대 엔트로피 모델(maximum entropy boosting model)을 제시하여 문서 단위화(text chunking)에 성공적으로 적용하였다. 최대 엔트로피 부스팅 모델은 쉬운 모델링과 높은 성능을 보이는 장점을 가지고 있다. 본 논문에서는 최대 엔트로피 부스팅 모델을 영어 전치사 접속 모호성 해소에 적용한다. Wall Street Journal 말뭉치에 대한 실험 결과, 아주 작은 노력을 들였음에도 84.3%의 성능을 보여 지금까지 알려진 최고의 성능과 비슷한 결과를 보였다.

• Language and Information
• /
• v.8 no.1
• /
• pp.101-116
• /
• 2004

This paper proposes a statistical model for the translation of prepositions in English-Korean machine translation. In the proposed model, statistical information acquired from unlabeled Korean corpora is used to choose the best translation from several possible translations. Such information includes functional word-verb co-occurrence information, functional word-verb distance information, and noun-postposition co-occurrence information. The model was evaluated with 443 sentences, each of which has a prepositional phrase, and we attained 71.3% accuracy.

• Cross-Cultural Studies
• /
• v.27
• /
• pp.387-414
• /
• 2012

This paper is intended to describe the idiosyncratic aspects of the 'let alone' construction in English, clarifying the productivity property of this construction and to show that this construction displays the properties of prepositional phrase that denote the meaning of 'the comparison of degree', against the claim that this construction should be regarded as a coordinating conjunction. In the process of doing that, this paper classifies this 'let alone' construction into two types of construction, by investigating the external and internal syntactic distributions of 'let alone' constructions as well as the relationship between 'let alone' phrases and the main clause.

• English Language & Literature Teaching
• /
• v.16 no.1
• /
• pp.41-59
• /
• 2009

The present study investigated Korean learners' knowledge on the range of possible interpretations of English locative PPs with manner of motion verbs, and considers whether learners can arrive at a superset L2 grammar on the basis of positive L2 input. Unlike Korean, some English locative PPs occurring with manner of motion verbs (such as in John jumped on the bed) are ambiguous as they can be interpreted as either directional or locational. Thirty Korean learners of English in three distinct groups (Advanced EFL-only group; Intermediate-EFL-only group; and ESL-experienced group) participated in an experimental study, along with a control group of nine native speakers of English. The results of the study showed that I) Korean learners, overall, tended to interpret English locative PPs as only locational, failing to recognize the ambiguity between the directional and locational readings in the target structure; 2) For the learners who experienced only the EFL context, even highly proficient learners, as well as intermediate level learners, failed to acknowledge the ambiguity; 3) The learners who experienced the ESL context for an extended period of time could identify the target reading to some extent, although they still could not reach the native-like competence. From these results, it is argued that robustness of positive evidence, not simply its availability, is critical in the acquisition of the superset L2 targets like the present one.

• The Journal of the Korea Contents Association
• /
• v.20 no.7
• /
• pp.274-281
• /
• 2020

N-gram analysis casts a new look at the n-word cluster in use different from the previously known idioms. It analyzes a corpus of English textbooks for frequently occurring n consecutive words mechanically using a concordance software, which is different from the previously known idioms. The current paper aims at extracting and comparing 4-gram words clusters between South Korean high school English textbooks and its North Korean counterpart. The classification criteria includes number of tokens and types between the two across oral and written languages in the textbooks. The criteria also use the grammatical categories and functional categories to classify and compare the 4-gram words clusters. The grammatical categories include noun phrases, verb phrases, prepositional phrases, partial clauses and others. The functional categories include deictic function, text organizers, stance and others. The findings are: South Korean high school English textbook contains more tokens and types in both oral and written languages. Verb phrase and partial clause 4-grams are grammatically most frequently encountered categories across both South and North Korean high school English textbooks. Stance is most dominant functional category in both South and North Korean English textbooks.