• Journal of Korea Multimedia Society
• /
• v.17 no.11
• /
• pp.1286-1295
• /
• 2014

This paper suggests the way that it could improve the reliability about preference of user's feedback by adding weighting factor on sentiment analysis, and efficiently make a sentiment analysis of users' emotional perspective on the big data massively generated on twitter. To solve errors on earlier studies, this paper has improved recall and precision of sensibility determination by using sensibility dictionary subdivided sentiment polarity based on the level of sensibility and given impotance to sensibility determination by populating slang, new words, emoticons and idiomatic expressions not in the system dictionary. It has considered the context through conjunctive adverbs fixed in korean characteristics which are free to the word order. It also recognize sensibility words such as TF(Term Frequency), RT(Retweet), Follower which are weighting factors of preference and has increased reliability of preference analysis considering weight on 'a very emotional tweet', 'a recognised tweet from users' and 'a tweeter influencer'

• Journal of the Korean Society of Safety
• /
• v.30 no.4
• /
• pp.151-158
• /
• 2015

Usually risk assessment is performed for the safety of diverse industries though, many kinds of risks cannot be analyzed effectively by using classical probability models due to lack of experience data and impreciseness of human decision making. For these reasons, fuzzy risk assessment utilizing subjective judgment and experience of skillful experts has been considered as a solution. In this study, to comprehend the relationship between conventional fuzzy theory and human conceptual images on risks, linguistic variables were reviewed with reference to fuzzy membership functions, especially in the Korean language. As interviewees, about a hundred people including students as well as safety engineers voluntarily participated. The research results showed that most people were in favor of adjective expressions decorated with adverbs rather than naive expressions such as "high" or "low", and that directly translated linguistic variables were not appropriate for the Korean people in risk assessment as far. Therefore, with consideration of the selection tendency by the Korean people in linguistic variables, it could be concluded that 5 level expressions would be most favorable for linguistic variables in risk assessments in Korea.

• Language Facts and Perspectives
• /
• v.48
• /
• pp.195-227
• /
• 2019

Building upon the state-of-the-art deep learning techniques, the present study classifies the texts written by Korean EFL learners and English native speakers and thereby demonstrates how the two types of texts differ from each other. To this end, the current work makes use of the Yonsei English Learner Corpus (YELC) and Gacheon Learner Corpus (GLC) as the L2 data, and Corpus of Contemporary American English (COCA) as the L1 data. Utilizing the sentence classification methods, the current work implements a system to differentiate the two types of texts, the accuracy of which is about 94%. This indicates that the deep leaning-based system is capable of identifying the well-formedness and felicities of the texts written by Korean EFL learners. Nonetheless, the system-based judgments do not overlap with human judgments largely because the deep learning model exclusively focuses on sequence of words. The present study provides a further analysis to see how the two types of judgments differ with respect to grammatical errors (e.g., word order, voice, etc.) and felicity errors (e.g., semantic prosody, the position of adverbs, etc.).

• English Language & Literature Teaching
• /
• v.18 no.1
• /
• pp.197-216
• /
• 2012

This study examines spoken interlanguage of Korean learners of English, focusing on the distribution of modal verbs and devices of epistemic modality. (Semi-) spontaneous speech data were collected from four students participating in a self-organized study group for seven months, which produced a corpus of about 55,000 words. The data analysis reveals the following: 1) The frequency of the modal verbs produced by the learners was lower than that of native speakers; 1.99 vs. 2.32 tokens per 100 words. The range of the modal verbs used by the learners was also very limited, with over-reliance on can (43%). 2) The grammatical categories of the devices marking epistemic modality were in the order of adverbs, lexical verbs, and modal verbs, with a high frequency of a few items in each category. 3) Lexical items conveying certainty and modals of obligation were preferred over markers of weaker commitment, resulting in speech characterized by firmer assertions and a more authoritative tone, a potential cause for pragmatic failure. 4) A weak developmental change was observed in the frequency of modal verbs, but not in their functions over the seven month period of data collection. L1 influence, L2 proficiency, mode of communication, and instruction effects are discussed as possible variables involved in the distribution patterns observed.

• Speech Sciences
• /
• v.10 no.3
• /
• pp.175-185
• /
• 2003

Japanese and Turkish are syntactically similar to each other, and there are syntactic structures that become ambiguous in terms of NP's and VP's modified by adjectives and adverbs, respectively. The prosody can play a crucial role in differentiating ambiguous sentences to correctly reflect their intended syntactic structures. In what way do the speakers of Sendai dialect of Japanese and Turkish use prosodic elements to differentiate syntactic ambiguities? Acoustic measurements were made of utterances of ambiguous sentences in Japanese and Turkish to observe prosodic strategies for disambiguation. Materials were sentences of the type ADV-VP1-NP-VP2, ADV-NP1-NP2-VP2, where the ambiguity lies in locative adverbial modification, ADV modifying either VP1 or VP2. For this construction the Japanese and Turkish creates the same ambiguities. In this paper, I look at duration, F0, and pause as observed in their speech and see how each language uses three prosodic elements in disambiguation. The results show that both speakers of Sendai dialect in Japan and those of Ankara dialect in Turky use lengthening of the syllable and/or pause before the boundary.

• Journal of Korean language education
• /
• v.28 no.3
• /
• pp.53-83
• /
• 2017

The goal of this study is to examine whether the genre characteristics of newspaper articles are appropriately reflected in Korean language textbooks. For the purpose of this study, two corpora were built with 17 textbook articles and 60 newspaper articles respectively. The average sentence length and frequency of vocabulary in each corpus were measured. It was found that the sentences of articles in textbooks tended to have longer sentence length and more complicated structures than the articles in newspapers. For instance, sentences in the textbook articles had more verbal endings, such as conjunctive and transforming endings. On the other hand, in case of vocabulary representing 'timeliness', there was a high frequency of adverbs and nouns which were related to year, month, and time in actual articles, while it is found to be very limited in textbooks. Also, typical translative styles such as '-ko itta', '-e ttareumyun' were more prominent in textbooks than in newspaper articles. In the case of abbreviated and omitted form of particles, this was a characteristic that appeared only in actual articles because of the constraint of space. It is significant that this paper offers suggestions for the development of reading materials for Korean language education by revealing that the genre typology of actual newspaper articles is not adequately reflected in current textbooks.

• Journal of KIISE:Software and Applications
• /
• v.36 no.1
• /
• pp.92-108
• /
• 2009

The Princeton WordNet (PWN), which was developed during last 20 years since the mid 80, aimed at representing a mental lexicon inside the human mind. Its potentiality, applicability and portability were more appreciated in the fields of NLP and KE than in cognitive psychology. The semantic and knowledge processing is indispensable in order to obtain useful information using human languages, in the CMC and HCI environment. The PWN is able to provide such NLP-based systems with 'concrete' semantic units and their network. Referenced to the PWN, about 50 wordnets of different languages were developed during last 10 years and they enable a variety of multilingual processing applications. This paper aims at describing PWN-referenced Korean Wordnet, KorLex 1.5, which was developed from 2004 to 2007, and which contains currently about 130,000 synsets and 150,000 word senses for nouns, verbs, adjectives, adverbs, and classifiers.

• Cross-Cultural Studies
• /
• v.38
• /
• pp.433-459
• /
• 2015

This paper aims at a better understanding of the concept of korean separable syntactic atom, of which inner parts is separable in syntax, and at examining whether or not this concept can apply to derivatives, functional complex constructions, idiomatic expressions in korean. I defined a syntactic atom as a minimum unit which is drawn directly from lexicon and then is applied to syntactic rules. And I insist that so-called 'lexical island constraint' has some problems and that the syntactic rules can be applied to inner parts of syntactic atom, if the syntactic rules is irrelevant to new syntactic atom formation. The greater part of derivatives is non-separable syntactic atoms. But the likes of '반짝거리다', '죄송스럽다', '칭얼대다' are the separable syntactic atoms. The degree of separability of them is different in the insertion of korean particles or negative adverbs and the omission of root of sytactic atom. The derivatives of 'X-적', of which roots is regular nominal roots, permit the syntactic link between roots and the syntactic combination of the root and its argument. These kinds of derivatives is separable syntactic atoms. Also the derivatives of 'bracket paradox' and 'X-답-' derivatives is separable syntactic atoms. All functional complex constructions are not separable syntactic atoms. According to the degree of grammaticalization, inner parts of some are separable, some is non-separable. Separable functional complex constructions only permit the switching of endings or Josas but not application of other syntactic rules. All idiomatic expressions which are composed of two or more syntactic atoms are separable syntactic atoms. Some of them have so strong separability to allow the insertion of syntactic atom, adverb or adnominal modification and the noun in idiomatic expression to become the head of the relative clause. And some idiomatic expressions which have weak separability only permit interrogative's substitution or form change in fraction of idiomatic expressions.

• Journal of the Korea Computer Industry Society
• /
• v.2 no.12
• /
• pp.1581-1590
• /
• 2001

This paper asserted more developmental model of WSD(word sense disambiguation) than J. Hur(2000)'s WSD model. This model suggested an improved statistical homonym disambiguation Model based on Bayes Theory. This paper using semantic information(co-occurrence data) obtained from definitions of part of speech(POS) tagged UMRD-S(Ulsan university Machine Readable Dictionary(Semantic Tagged)). we extracted semantic features in the context as nouns, predicates and adverbs from the definitions in the korean dictionary. In this research, we make an experiment with the accuracy of WSD system about major nine homonym nouns and new seven homonym predicates supplementary. The inner experimental result showed average accuracy of 98.32% with regard to the most Nine homonym nouns and 99.53% for the Seven homonym predicates. An Addition, we save test on Korean Information Base and ETRI's POS tagged corpus. This external experimental result showed average accuracy of 84.42% with regard to the most Nine nouns over unsupervised learning sentences from Korean Information Base and ETRI Corpus, 70.81 ％ accuracy rate for the Seven predicates from Sejong Project phrase part tagging corpus (3.5 million phrases) too.

• KIISE Transactions on Computing Practices
• /
• v.21 no.6
• /
• pp.412-417
• /
• 2015

Context-sensitive spelling errors (CSSE) are hard to correct, since they are perfect words when analyzed alone. Determined only by considering the semantic and syntactic relations of their context, CSSEs affect largely the performance of spelling and grammar checkers. The existing Korean Spelling and Grammar Checker (KSGC 4.5) adopts a rule-based method, which uses hand-made correction rules for CSSEs. Using rule-based method, the KSGC 4.5 is designed to obtain the very high precision, which results in the extremely low recall. In this paper, we integrate our previous works that control the CSSE correction rules, in order to improve the recall without sacrificing the precision. In addition to the integration, facultative insertion of adverbs and conjugation suffix of predicates are also considered, as for constraint-loosening linguistic features.