• 지식경영연구
• /
• 제23권2호
• /
• pp.253-276
• /
• 2022

온라인 리뷰는 소비자의 구매 의사결정에 중요한 역할을 하기 때문에 소비자의 지각된 리뷰 유용성에 영향을 미치는 리뷰 요인을 확인하는 것이 필요하다. 그러나 온라인 리뷰의 유용성에 대한 대부분의 기존 연구는 주로 리뷰 및 리뷰어 속성과 같은 정량적 요인에 중점을 두고 있다. 최근 연구에서는 리뷰 콘텐츠과 언어 스타일이 소비자의 구매 의사결정에 미치는 영향을 조사했다. 또한, 소비자가 의사결정 과정에서 리뷰를 평가할 때 추가적으로 리뷰 텍스트 속성들을 고려해야 한다고 주장하고 있다. 따라서 본 연구는 온라인 리뷰 맥락에서 리뷰 콘텐츠과 언어 스타일이 리뷰 유용성에 미치는 영향을 조사하고자 한다. 추가적으로 리뷰어의 전문성이 리뷰 콘텐츠 및 언어 스타일과 리뷰 유용성 간의 영향관계를 조절하는지 여부를 조사했다. 연구결과 긍정적인 리뷰 콘텐츠는 리뷰 유용성에 부정적인 영향을 미치고, 부정적인 리뷰 콘텐츠와 언어적 스타일은 리뷰 유용성에 긍정적인 영향을 미치는 것으로 나타났다. 리뷰어의 전문성은 부정적인 리뷰 콘텐츠와 언어 스타일이 리뷰 유용성에 미치는 영향을 완화시키는 것으로 나타났다. 본 연구결과는 온라인 전자상거래 기업이 리뷰 유용성에 영향을 미치는 요인을 파악하고 이를 마케팅 전략에 반영하여 기업 매출 성장을 달성하는데 시사점을 제공할 수 있다.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제11권3호
• /
• pp.125-132
• /
• 2022

문장 압축은 원본 문장의 중요한 의미는 유지하면서 길이가 축소된 압축 문장을 생성하는 자연어처리 태스크이다. 문법적으로 적절한 문장 압축을 위해, 초기 연구들은 사람이 정의한 언어 규칙을 활용하였다. 또한 시퀀스-투-시퀀스 모델이 기계 번역과 같은 다양한 자연어처리 태스크에서 좋은 성능을 보이면서, 이를 문장 압축에 활용하고자 하는 연구들도 존재했다. 하지만 언어 규칙을 활용하는 연구의 경우 모든 언어 규칙을 정의하는 데에 큰 비용이 들고, 시퀀스-투-시퀀스 모델 기반 연구의 경우 학습을 위해 대량의 데이터셋이 필요하다는 문제점이 존재한다. 이를 해결할 수 있는 방법으로 사전 학습된 언어 모델인 BERT를 활용하는 문장 압축 모델인 Deleter가 제안되었다. Deleter는 BERT를 통해 계산된 perplexity를 활용하여 문장을 압축하기 때문에 문장 압축 규칙과 모델 학습을 위한 데이터셋이 필요하지 않다는 장점이 있다. 하지만 Deleter는 perplexity만을 고려하여 문장을 압축하기 때문에, 문장에 속한 단어들의 언어 정보를 반영하여 문장을 압축하지 못한다. 또한, perplexity 측정을 위한 BERT의 사전 학습에 사용된 데이터가 압축 문장과 거리가 있어, 이를 통해 측정된 perplexity가 잘못된 문장 압축을 유도할 수 있다는 문제점이 있다. 이를 해결하기 위해 본 논문은 언어 정보의 중요도를 수치화하여 perplexity 기반의 문장 점수 계산에 반영하는 방법을 제안한다. 또한 고유명사가 자주 포함되어 있으며, 불필요한 수식어가 생략되는 경우가 많은 뉴스 기사 말뭉치로 BERT를 fine-tuning하여 문장 압축에 적절한 perplexity를 측정할 수 있도록 하였다. 영어 및 한국어 데이터에 대한 성능 평가를 위해 본 논문에서 제안하는 LI-Deleter와 비교 모델의 문장 압축 성능을 비교 실험을 진행하였고, 높은 문장 압축 성능을 보임을 확인하였다.

• 인문언어
• /
• 제7집
• /
• pp.57-92
• /
• 2005

The main driving force behind Chomskyan theory of language that is gaining ground under the title of the "minimalist program" has been the urge to reduce the variety of natural languages to a minimal number of theoretical concepts and devices and to formulate the strongest possible hypothesis about linguistic structures. This has led to a long series of proposals of ever greater abstractness, with concomitant modifications in a number of theoretical constructs. The minimalist approach is the latest development of these constant changes, for which there is really "only one computational system and one lexicon," the differences among languages being reducible to parametric variation in "nonsubstantive parts of the lexicon"(Chomaky 1995:169-70). Chomsky thus differs, now more than ever, from other linguists by his sweepingly programmatic, rather than empirical, approach to language. The proposals he makes are too complicated to discuss here in any technical details. We rather focus on a series of lectures delivered by an ardent adherent of the minimalist program, with the view to demonstrating how this theory may mislead and distort the whole prospect of linguistic investigation. The rationale of the so-called 'minimalism' per se will be called into question.

• International Journal of Contents
• /
• 제6권2호
• /
• pp.33-40
• /
• 2010

This study aims at investigating the linguistic and educational factors affecting TOEFL scores, focusing on three OECD countries, Korea, Japan, and Finland. The data comprise document analysis on curriculums, websites, and literature. The findings reveal that the number of Korean test-takers and their TOEFL scores gradually increased year by year. Finnish test-takers consistently gained greatly high scores, and Japanese examinees showed the lowest scores. The languages Korean, Japanese, and Finnish are all far distant from English and receive little support on historical grounds from the Indo-European family tree. In Finland, however, Swedish which belongs to Indo-European languages is still used as an official language with Finnish. Korea and Finland adopt English education from Year 3 in primary school, whereas English is not an official subject in primary school at present in Japan. Finnish students are taught a foreign language in addition to English from primary school. These seem to support the result of the high TOEFL scores of Finnish test-takers. This study concludes that social context which includes linguistic and educational environments are the main factors which affect TOEFL scores.

• 영어어문교육
• /
• 제6호
• /
• pp.89-111
• /
• 2000

Since speaking is correlated with time and time is metaphorically conceptualized in terms of space, it is natural for us to conceptualize language metaphorically in terms of space. For example, we think that the future is in front and the past is behind. Reddy(1979) suggested in his conduit metaphor that linguistic expressions are containers. According to him, the speaker puts his ideas(objects) into words{containers)and sends them along a conduit to a hearer who takes the idea(object) out of the word(container). As a result we are able to know that the larger linguistic expressions have more meaning in it. In other words the space of a linguistic form has close relationship with meaning. Moreover we are able to see that formal distance between arguments or elements of linguistic expressions shows semantic influences between them. If two elements keep close distance, a preceeding element has a strong, direct and whole influence upon the following element. Sometimes even the results of the influence can be brought out implicitly by the formal relation between two elements. Therefore, the purpose of this paper is to show that tins difference in meaning which is due to formal distance of sentence elements can be explained by the metaphorical concept presented by Lakoff and Johnson(1980).

• 영어어문교육
• /
• 제1호
• /
• pp.69-108
• /
• 1995

Listening comprehension can be defined as a process of an integrative, positive and creative activity through which listeners get the message of speakers' production using linguistic or non-linguistic redundancy as well as linguistic or non-linguistic knowledge. Compared with reading comprehension, it has many difficulties especially for foreigners. while it can be transferred to the other skills: speaking, reading, writing. With this said, listening comprehension can be taught effectively using the following teaching strategies. First. systematic and intensive instruction of segmental phonemes, suprasegmental phonemes and sound changes must be given to remove the difficulties of listening comprehension concerned with the identification of sounds. Second, vocabulary drill through various games and other activities is absolutely needed until words can be unconsciously recognized. Without this, comprehension is almost impossible. Third, instruction of sentence structures is thought to be essential considering grammar is supplementary to listening comprehension and reading comprehension for academic purpose. So grammar translation drills, mechanical drills, meaningful drills and communicative drills should be performed in succession with common or frequently used structures. Fourth, listening activities for overall comprehension should teach how to receive overall meaning of intended messages intact. Linguists and literatures have listed some specific activities as follows: Total Physical Response, dictation, role playing, singing songs, selective listening, picture recognition, list activities, completion, prediction, true or false choice, multiple choice, seeking of specific information, summarizing, problem-solving and decision-making, recognization of relationships between speakers, recognition of mood, attitude and behavior of speakers.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제11권3호
• /
• pp.165-170
• /
• 2011

Due to advances in hardware performance, user-friendly interfaces are becoming one of the major concerns in information systems. Linguistic conversation is a very natural way of human communications. Fuzzy techniques have been employed to liaison the discrepancy between the qualitative linguistic terms and quantitative computerized data. This paper deals with linguistic queries using clustering results on data sets, which are intended to retrieve the close clusters or distant clusters from the clustering results. In order to support such queries, a fuzzy technique-based method is proposed. The method introduces distance membership functions, namely, close and distant membership functions which transform the metric distance between two objects into the degree of closeness or farness, respectively. In order to measure the degree of closeness or farness between two clusters, both cluster closeness measure and cluster farness measure which incorporate distance membership function and cluster memberships are considered. For the flexibility of clustering, fuzzy clusters are assumed to be formed. This allows us to linguistically query close or distant clusters by constructing fuzzy relation based on the measures.

• 한국언어정보학회지:언어와정보
• /
• 제17권2호
• /
• pp.93-121
• /
• 2013

Identifying the speakers of individual utterances mentioned in textual stories is an important step towards developing applications that involve the use of unique characteristics of speakers in stories, such as robot storytelling and story-to-scene generation. Despite the usefulness, it is a challenging task because not only human entities but also animals and even inanimate objects can become speakers especially in fairytales so that the number of candidates is much more than that in other types of text. In addition, since the action of speaking is not always mentioned explicitly, it is necessary to infer the speaker from the implicitly mentioned speaking behaviors such as appearances or emotional expressions. In this paper, we investigate a method to exploit linguistic clues to identify the speakers of utterances from textual fairytale stories in Korean, especially in order to handle such challenging issues. Compared with the previous work, the present work takes into account additional linguistic features such as vocative roles and pairs of conversation participants, and proposes the use of discourse-level turn-taking behaviors between speakers to further reduce the number of possible candidate speakers. We describe a simple rule-based method to choose a speaker from candidates based on such linguistic features and turn-taking behaviors.

• Journal of Electrical Engineering and Technology
• /
• 제11권2호
• /
• pp.309-319
• /
• 2016

This paper presents a new recognition and classification method for power quality (PQ) disturbances on the basis of pattern linguistic values. This method solves the difficulty of recognizing disturbances rapidly and accurately by using fuzzy logic. This method uses classification disturbance patterns to define the linguistic values of fuzzy input variables and used the input variables of corresponding disturbance pattern to set membership functions. This method also sets the fuzzy rules by analyzing the distribution regularities of the input variable values. One characteristic of this method is that the linguistic values of fuzzy input variables and the setting of membership functions are not only related to the input variables but also to the character of classification disturbance and the classification results. Furthermore, the number of fuzzy rules is equal to the number of disturbance patterns. By using this method for disturbance classification, the membership function and design of fuzzy rules are directly related to the objective of classification, thus effectively reducing the complexity of the design process and yielding accurate classification results. The classification results of the simulation and measured data verify the feasibility and effectiveness of this method.

• 한국의상디자인학회지
• /
• 제16권1호
• /
• pp.91-103
• /
• 2014

In this study, 70 national brands among men's wear brands were selected to examine linguistic characteristics of domestic national men's wear brand names. Linguistic factors which were used in national men's wear brand names were analyzed to understand their characteristics. Formative and semantic characteristics of each brand name were analyzed on the basis of the results from previous studies. It was found that long words with over four syllables are preferred than short words and single words in the form of noun are frequently used for domestic national men's wear brand names in terms of linguistic formality. English is most widely used in brand names, and European languages such as French, Spanish, and Italian are also used frequently under the influence of the country of origin. Next, the analysis result on the semantic characteristics of domestic national men's wear brand names showed that descriptive brand names are used to convey brand information directly and easily, or freestanding brand names which are absolutely irrelevant and newly coined words are chosen to create a characteristic image. In other words, brand names represent detailed business and product category of men's wear by forming a brand image of men's wear (ex. Man, Homme, Zio), and provide the information about properties and benefits related to the product such as dignity, masterpiece, and luxurious lifestyle to consumers by presenting the concept of the brand.