• Journal of Internet Computing and Services
• /
• v.16 no.2
• /
• pp.41-48
• /
• 2015

On movie review sites, movie ratings are determined by netizens' subjective judgement. This means that inconsistency between ratings and opinions from netizens often occurs. To solve this problem, this paper proposes sentiment sentence sets which affect movie evaluation, and apply sets to comments to infer ratings. Creation of sentiment sentence sets is consisted of two stages, construction of sentiment word dictionary and creation of sentiment sentences for sentiment estimation. Sentiment word dictionary contains sentimental words and its polarities included in reviews. Elements of sentiment sentences are combined with movie related noun and predicate from words sentiment word dictionary. In this study, to make correspondence between polarity of sentiment sentence and sentiment word dictionary, sentiment sentences which have different polarity with sentiment word dictionary are removed. The scores of comments are calculated by applying averages of sentiment sentences elements. The result of experiment shows that sentence scores from sentiment sentence sets are closer to reflect real opinion of comments than ratings by netizens'.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.5 s.43
• /
• pp.75-86
• /
• 2006

Target word selection is one of the most important and difficult tasks in English-Korean Machine Translation. It effects on the translation accuracy of machine translation systems. In this paper, we present a new approach to select Korean target word for an English noun with translation ambiguities using multiple knowledge such as verb frame patterns, sense vectors based on collocations, statistical Korean local context information and co-occurring POS information. Verb frame patterns constructed with dictionary and corpus play an important role in resolving the sparseness problem of collocation data. Sense vectors are a set of collocation data when an English word having target selection ambiguities is to be translated to specific Korean target word. Statistical Korean local context Information is an N-gram information generated using Korean corpus. The co-occurring POS information is a statistically significant POS clue which appears with ambiguous word. The experiment showed promising results for diverse sentences from web documents.

• Cross-Cultural Studies
• /
• v.45
• /
• pp.325-348
• /
• 2016

This study aims to identify the representation of migrants in the English-language media during the Brexit campaign period. For the purpose of this study, the methodological tool of corpus-assisted discourse studies(CADS) was employed and a collection was compiled of articles mentioning Brexit in British, American, Canadian, and Australian media from April 15 to June 22, 2016 in order to compare their portrayals of migrants. To examine how IMMIGRANT, MIGRANT, and REFUGEE are represented in the media, their collocates were analyzed by MI score and categorized by social actor categorization(Van Leeuwan, 1996). The results show that IMMIGRANT is related to collocates that refer to legal status and provenance, MIGRANT associated with economic terms, and REFUGEE relates to terms expressing quantities. The results also reveal that migrants are frequently depicted by functionalization, classification, and appraisement categorization and are more negatively portrayed in British and American media. This paper claims that corpus-assisted linguistic analysis of words enables one to identify salient linguistic patterns or lexical choices in the discourses about a particular phenomenon or group of people.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1998.08a
• /
• pp.56-59
• /
• 1998

POW 3848 DB 및 SNR 이 크게 다른 2 종류의 PC168 DB를 대상으로 가변어휘 음성인식 시스템을 이용하여 훈련 및 성능 평가 실험을 수행한 내용에 대해서 기술하고 있다. 실험의 목적은 위의 3종류의 DB를 조합하여 얻은 DB 환경하에서 인식기를 훈련시키면서, DB 의 조합 및 훈련방법에 따른 인식기의 성능과의 상관관계를 도출하고자 하였다. DB 의 조합은 POW DB 와 SNR 이 높은 PC DB , 및 3종류의 DB 모두로 구성하였다. 인식기는 40개의 음소로 구성된 문맥 독립형 SCHMM 모델이며, 각 음소당 3개의 상태로 이루어져 있다. 실험 결과, 대부분의 경우에서 ITERATION이 1.0인 경우에 최고 인식률을 나타내고 있으며, INTERATION 이 3.0 이상인 경우에는 항상 CASE 3의 실험방법이 우세한 결과를 나타내었다. 또한 CASE 1으로 훈련한 경우가 CASE 2 보다는 각각의 실험 DB 에 대해서 대체적으로 좋은 결과를 보였다.

• Proceedings of the Korean Society for Emotion and Sensibility Conference
• /
• 1997.11a
• /
• pp.163-167
• /
• 1997

인간이 움직이는 물체에 탑승하고 있거나 움직이는 물체를 잡고 있을때 어떤 종류의 감성을 느끼게 된다. 물체의 속도, 가속도, 또는 강성, 감쇠 등으로 인하여 인간은 쾌, 불쾌감을 갖게 되며, 본 논문에서는 이러한 감성을 운동감이라 정의한다. 이러한 운동감을 공학적으로 유용한 데이터로 만들기 위해 운동감 어휘를 도입하여 정량화를 시도하였으며, 복수의 운동감 어휘를 연산할 수 있는 방법과 가중치를 구할 수 있는 방법을 제시하고자 한다. 본 연구에서는 귀의 전정 기관에서 느끼는 몸 전체의 평형 감각 및 운동 감각은 고려의 대상으로 제외하며, 팔에 국한하여 피부 감각과 팔 근육의 위치 인지 등으로 인한 운동감을 해석 대상으로 한다. 해석의 편의성을 위하여 팔을 제외한 몸의 움직임은 없는 상태로 유지하며, 팔의 2차원 운동만을 고려하기고 한다. 퍼지는 사람의 언어와 같이 모호한 사건을 해석하기 위한 이론이다. 모호한 정도를 표현하는 방법으로 퍼지 정도척도(measutr of fuzziness)와 퍼지척도(fuzzy measure)가 많은 분야에서 이용되고 있다. 하지만 운동감에 대한 연구는 미비한 실정이므로 불확실성을 평가하는 퍼지 이론을 이용하여 운동감을 해석하려 한다.

• Annual Conference of KIPS
• /
• 2013.05a
• /
• pp.257-259
• /
• 2013

상품평은 구매 예정자의 의사 결정에 큰 도움을 준다. 그러나 하나의 상품에 관한 상품평의 수가 많고 의견도 다양하여 모든 상품평을 읽고 상품에 대한 직관적인 판단을 내리기가 어렵다. 본 논문에서는 하나의 상품에 대한 모든 상품평을 분석하고 각각의 속성별로 극성(긍정, 부정) 정보를 추출하여 구매 예정자에게 제공함으로써 해당 상품이 어떠한 평가를 받고 있는지 빠른 판단이 가능하게 한다. 한국어의 언어적 특징을 반영하여 속성별 어휘 자질을 추출하고 이를 바탕으로 상품의 속성별 극성을 판단한다. 또한, 기구축한 속성별 어휘 사전의 자료부족 문제로 말미암아 상품평을 분석할 수 없을 때는 전체 어휘의 극성정보를 이용하여 상품의 전체 극성을 판단한다.

• Korean Journal of Cognitive Science
• /
• v.13 no.3
• /
• pp.23-31
• /
• 2002

Previous studies on color sensibility employed mostly general emotional and/or sensational words, and could not have a focus on the matter of 'color'. This study tried to collect sensibility words of color widely, and select representative words in a systematic way, with a result of 90 sensibility word list. 25 participants were asked to indicate how they feel on each item of the list when each of 22 colored light was projected on the screen. Factor analysis of data extracted 5 sensibility factors, such as mildness, neatness, splendour, dynamic, and weightiness. They were thought to represent color sensibilities more specifically, while having some similarities to, than those factors proposed previously. Issues on interpretation of factors and application of this result were discussed.

• Journal of Hospice and Palliative Care
• /
• v.6 no.1
• /
• pp.11-21
• /
• 2003

This paper addresses the minor differences in the description of pain in Korean language in order to develop a standarized cancer pain aneument tool for Korean adults, Korean Cancer Pain Assessement Tool. The subtle differences in the meaning of expressions used cannot be translated into English and therefore we omiltted the English abstract.

• Korean Institute of Interior Design Journal
• /
• v.21 no.1
• /
• pp.220-227
• /
• 2012

After The Ramsar Convention took place in Korea in 2008, public interest in environment heightened, and the government has been allocating its budget for conserving our environment. Therefore, the present study focuses on eco-friendly spatial images particularly shown in ecological museums in Korea which recognizes the value of the environment and the ecology and tries to both protect them and alert people about it. The purpose of this study is proving what consist of eco-friendly spatial images by analyzing the expressive word of eco-friendly images and the image of space, providing a basic data for future space planning of ecological museums. To do this, the present study proceeds in three steps. First of all, the base of research in analyzing stage is firmly established by grasping general theories and terms regarding spatial image. As a second step, the composition and the characteristics of exhibition is clarified through on-spot investigation to provide comparative data for spatial image assessment in the future. Also through this step, we could understand how the exhibits are designed currently. In the last stage of research, expressive words regarding eco-friendly spatial images are extracted and used to analyze the spatial image of ecological museums. And the following three conclusions is deduced. First, the expressive words of eco-friendly spatial image that are extracted are as following: "healthy", "coexisting", "clean", "blending", "warm", "soft", "lively", "pure", "cool", "fresh ", "comfortable", "relaxed", "mild", "free", "harmonious", and "healing". As the second conclusion, color, and material, the formation which is an architectural factor did not have a great impact on forming eco-friendly image, but the color and the material did. The third conclusion was that the display with natural aspects actively utilized increased eco-friendly spatial image.

• Science of Emotion and Sensibility
• /
• v.10 no.3
• /
• pp.403-410
• /
• 2007

The most common method in the psychological measuring method for measuring user's emotions is to indirectly measure the user's emotion by using adjectives, called emotional words. The previous research, in order to observe user's emotional changes while they interact with a product, has extracted some emotional words and representative emotions, and made a set of subjective evaluation scale. In addition to adjective checklists, self-report questionnaire-type scales have been extensively used to assess user's emotions. This research suggested a self-report questionnaire-type scale using the representative emotions and a set of subjective evaluation scale made in the previous research. Also the reliability of the suggested self-report questionnaire-type scale was confirmed through the analysis of Cronbach's coefficient alpha. Therefore, the self-report questionnaire-type scale extracted through this research can be used in various ways to measure a user's user's emotions naturally expressed while using a product.