• 한국멀티미디어학회논문지
• /
• 제18권11호
• /
• pp.1391-1399
• /
• 2015

In recent times, studies sentiment analysis are being actively conducted by implementing natural language processing technologies for analyzing subjective data such as opinions and attitudes of users expressed on the Web, blogs, and social networking services (SNSs). Conventionally, to classify the sentiments in texts, most studies determine positive/negative/neutral sentiments by assigning polarity values for sentiment vocabulary using sentiment lexicons. However, in this study, sentiments are classified based on Thayer's model, which is psychologically defined, unlike the polarity classification used in opinion mining. In this paper, as a method for classifying the sentiments, sentiment categories are proposed by extracting sentiment keywords for major sentiments by using hashtags, which are essential elements of Instagram. By applying sentiment categories to user posts, sentiments can be determined through the similarity measurement between the sentiment adjective candidates and the sentiment keywords. The test results of the proposed method show that the average accuracy rate for all the sentiment categories was 90.7%, which indicates good performance. If a sentiment classification system with a large capacity is prepared using the proposed method, then it is expected that sentiment analysis in various fields will be possible, such as for determining social phenomena through SNS.

• 복식
• /
• 제62권2호
• /
• pp.13-25
• /
• 2012

The purpose of this research is to study the influence over confidence and Word-Of-Mouth (WOM) effect in the acceptance process of WOM by information characteristics(consensus, vividness, and message neutrality) of individual cosmetic blog, reflecting that it is widely spreading over the public. Online survey for the consumers using such blogs was performed to collect data(N=200), and credibility analysis through Cronbach-${\alpha}$ and Structural Equation Model(SEM) analysis using AMOS 18.0 were performed. The analysis results are as follows; First, vivid information and neutral message increase WOM effect through the improvement of consumers' credibility whereas information consensus doesn't positively influence over credibility. Second, the examination of a moderating effect by type of cosmetics demonstrated that the consumers of basic cosmetics have most confidence in vivid information, and those of color cosmetics do more confidence in neutral message. Therefore, there is a difference in the credibility factor of online WOM depending on the type of cosmetics consumption so that differentiated information provision strategies for cosmetics groups should be established based on it.

• 정보관리학회지
• /
• 제27권1호
• /
• pp.61-77
• /
• 2010

최근 다양한 주제 분야의 블로그가 이용자의 정보요구를 충족시켜주는 웹 정보원 중 하나로 활용되고 있다. 본 연구에서는 블로그 페이지의 검색 성능을 향상시키기 위하여 이용자가 부여한 태그 및 트랙백을 이용하여 블로그 페이지의 검색 실험을 수행하였다. 실험을 위해 4,908개의 블로그 페이지와 각 페이지에 트랙백으로 연결된 다른 블로그 페이지의 URL을 수집하였다. 검색 자질로 본문의 용어에 이용자 태그를 추가하였을 경우와 네트워크 중심성 값을 반영하였을 경우 모두 검색 성능이 향상되었고, 본문 용어와 이용자 태그를 검색 자질로 함께 사용하고 여기에 중심성 값을 반영하였을 경우 가장 좋은 성능을 보였다.

• 한국콘텐츠학회:학술대회논문집
• /
• 한국콘텐츠학회 2007년도 추계 종합학술대회 논문집
• /
• pp.116-120
• /
• 2007

한민족과학기술자네트워크 KOSEN에서는 과학기술자 블로그 서비스를 시작하였다. 기존의 정형화된 서비스의 틀을 넘어 과학자 자신이 자유롭게 자신이 보유한 지식을 드러내며 자유로운 교류가 가능케 하기 위한 시도이다. 2007년 4월 처음 개설한 이래 2007년 10월 현재 251개의 블로그가 개설되어 운영중이다. 블로그를 통해 전공지식에서부터 여행, 취미까지 다양한 콘텐트들을 확보할 수 있었으며, 블로그 서비스의 문제점도 파악할 수 있었다. 본 논문에서는 과학자 블로그 사용 실태를 분석하여 제시한다. 또한 블로그 운영의 활성화 전략을 도출하여 서비스의 나아갈 방향을 제시하고자 한다.

• 한국산학기술학회논문지
• /
• 제11권8호
• /
• pp.3070-3075
• /
• 2010

웹 2.0 시대를 맞아 인터넷 상의 블로그 및 커뮤니티 공간에 일반 사용자들이 자신의 의견 및 생각을 표현하게 되었다. 상품 구매 시 다수의 사람들이 이러한 의견을 참조하는데, 사용자들은 소수의 의견만을 참조하고 전체적인 의견은 참조하지 못하고 있다. 의견 분석 시스템은 상품 및 서비스에 대한 인터넷 상의 글들을 분석하여 상품의 긍정, 부정을 평가하는 시스템으로 자연어 검색에서 발전한 검색이라 할 수 있다. 본 논문에서는 의견 분석 서비스에서 핵심이 되는 문장의 긍정, 부정을 파악하기 위하여 '긍정', '부정', '중립'의 극성 정보 외에 '반전'의 정보를 추가로 학습하고, 처리하는 구문 분석 및 반전 처리를 제안한다.

• Journal of Multimedia Information System
• /
• 제6권4호
• /
• pp.217-224
• /
• 2019

As the world's population grows, how to maintain the food supply is becoming a bigger problem. Now and in the future, big data will play a major role in decision making in the agriculture industry. The challenge is how to obtain valuable information to help us make future decisions. Big data helps us to see history clearer, to obtain hidden values, and make the right decisions for the government and farmers. To contribute to solving this challenge, we developed the Agriculture Big Data Analysis System. The system consists of agricultural big data collection, big data analysis, and big data visualization. First, we collected structured data like price, climate, yield, etc., and unstructured data, such as news, blogs, TV programs, etc. Using the data that we collected, we implement prediction algorithms like ARIMA, Decision Tree, LDA, and LSTM to show the results in data visualizations.

• International Journal of Contents
• /
• 제14권3호
• /
• pp.27-31
• /
• 2018

Social network service (SNS) data related to green tourism were used to estimate preferred tour sites and users' interests. Keywords related with green tour activities were employed to search the SNS data. SNS data were collected from Korean blogs such as Naver and Daum from June $1^{st}$ to August $31^{st}$ between 2015 and 2017 using text-mining solution. During the study period, seven hundred and five posts were analyzed. Associated words that frequently co-occurred with keywords were classified into different categories depending on the nature of associated words. Associated words included swimming pools and camping sites (location); experience and swimming pools (attribute); and water play and culture (culture/leisure). Our data suggest that SNS users with experience of green tourism in Korea exhibited interest in green tourism with swimming pools, camping sites, experience, water play and/or culture rather than particular popular sites. Based on the findings, it is recommended that preferred facilities such as swimming pools should be provided at green tourism sites to meet the users' needs and to facilitate green tourism.

• 정보처리학회논문지D
• /
• 제18D권2호
• /
• pp.101-110
• /
• 2011

블로그의 사용량이 증가함에 따라 다수의 포스트들이 블로고스피어 내에 작성되고 있으며, 이는 검색에서 웹 서퍼가 양질의 포스트를 찾기 어렵게 하는 문제를 가져왔다. 이로 인하여 포스트 검색에서 랭킹을 부여하기 위한 랭킹 알고리즘의 필요성이 부각되고 있다. 기존에 웹 문서를 위한 다양한 랭킹 알고리즘들이 있었으나, 웹 문서와 포스트의 차이로 인하여 직접 적용하기 어렵다는 문제점이 존재한다. 본 논문에서는 블로거들이 포스트에 남긴 블로그 액션을 이용하여 포스트에 랭킹을 부여하는 방안인 포스트 랭킹 알고리즘들을 제안한다. 그리고 실제 블로그 데이터를 이용하여 포스트 랭킹 알고리즘들의 성능을 분석하고, 이를 바탕으로 블로그에 적합한 포스트 랭킹 알고리즘을 선별한다.

• Journal of Information Technology Applications and Management
• /
• 제18권3호
• /
• pp.73-89
• /
• 2011

More and more employees are implementing the use of emerging Web 2.0 tools such as blogs, wikis, social networks, etc in workplaces. However, their attitudes towards adoption of Web 2.0 tools in workplaces still lack theoretical support. The purpose of this study aims to provide a conceptual examination of the determinants that influence the intention to use Web 2.0 applications in workplaces in Korea. To achieve this objective, this study selected the theory of reasoned action (TRA) as a theoretical basis to explain variation in behavioral intentions. Structural equation modeling was employed to analyze data collected from 269 workers distributed in 5 companies in Korea. In addition, we classified respondents into extroverts and introverts and delineated the different factors for these two types of respondents that affect their intentions to use Web 2.0 tools in workplaces. The findings of this research could provide a theoretical foundation for academics on the validation of technology adoption. This research will also serve as a guideline for service providers in designing the Web 2.0 services.

• International Journal of Internet, Broadcasting and Communication
• /
• 제12권2호
• /
• pp.113-119
• /
• 2020

Numerous tourist-related data produced on the Internet contain not only simple tourist information but also diverse ideas and opinions from users. In order to derive meaningful information about tourist sites from such big data, the social network analysis of tourist keywords can identify the frequency of keywords and the relationship between keywords. Thus, it is possible to make recommendations more suitable for users by utilizing the clear recommendation criteria of tourist attractions and the relationship between tourist attractions. In this paper, a recommendation system was designed based on tourist site information through big data social network analysis. Based on user personality information, the types of tourism suitable for users are classified through deep learning and the network analysis among tourist keywords is conducted to identify the relationship between tourist attractions belonging to the type of tourism. Tour information for related tourist attractions shown on SNS and blogs will be recommended through tagging.