Latent Dirichlet Allocation 기법을 활용한 해외건설시장 뉴스기사의 토픽 모델링(Topic Modeling)

문성현,정세환,지석호 대한토목학회 2018 대한토목학회논문집 Vol.38 No.4

해외건설 프로젝트를 기획하고 수행하는 과정에서 현지 시장의 상황을 신속하고 정확하게 파악하는 것은 수익성 창출에 매우 큰 영향을 미친다. 뉴스기사 데이터는 정치, 경제, 사회 등 다양한 관한 정보를 담고 있기 때문에 시장의 상황을 파악하는 데 사용할 수 있는 좋은 데이터이다. 텍스트의 형태로 존재하는 대량의 뉴스기사 데이터로부터 정보를 추출하고 내용을 요약하는 과정에서 인력, 비용, 시간의 소모를 줄이기 위해 텍스트마이닝 기술이 필요하다. 본 연구에서는 뉴스기사에 다양한 주제가 공존한다는 특성으로 인해 발생하는 정보 추출의 한계를 극복하기 위해 잠재 디리클레 할당(Latent Dirichlet Allocation) 방법론을 사용하여 토픽 모델링을 수행했다. 문서 집단에 존재하는 주제의 개수가 10개라고 가정했을 때, 이용자들의 편의 증진을 위한 프로젝트(2번 주제)와 아프리카 지역의 빈곤 문제를 해결하기 위한 민간 차원의 지원(4번 주제) 등의 주제 집단이 존재하는 것을 확인했다. 이와 같이 문서 집단의 주제를 구분함으로써 더욱 의미있는 정보를 추출하고, 요약 결과의 활용성을 높일 수 있다. Sufficient understanding of oversea construction market status is crucial to get profitability in the international construction project. Plenty of researchers have been considering the news article as a fine data source for figuring out the market condition, since the data includes market information such as political, economic, and social issue. Since the text data exists in unstructured format with huge size, various text-mining techniques were studied to reduce the unnecessary manpower, time, and cost to summarize the data. However, there are some limitations to extract the needed information from the news article because of the existence of various topics in the data. This research is aimed to overcome the problems and contribute to summarization of market status by performing topic modeling with Latent Dirichlet Allocation. With assuming that 10 topics existed in the corpus, the topics included projects for user convenience (topic-2), private supports to solve poverty problems in Africa (topic-4), and so on. By grouping the topics in the news articles, the results could improve extracting useful information and summarizing the market status.

Topic-wise Sentiment Analysis Based on LASSO regression and Latent Dirichlet Allocation: Focus on Customer Reviews on Hotel

갈리총,정욱 한국품질경영학회 2022 한국품질경영학회 학술대회 Vol.2022 No.1

In a situation where business competition is intensifying, companies can improve their services and increase their profits by understanding the topics most important to customers and their satisfaction with those topics. One way to do this is to get customer reviews for a particular product or service from an internet website. According to a survey, about 95% of travelers choose a hotel after reading hotel online reviews. As more and more people recognise the importance of hotel online reviews, more research is being conducted on topic mining. In this paper, LASSO regression analysis was used to build a word sentiment dictionary, and the topics of hotel reviews were modelled with Latent Dirichlet Allocation. Finally, topic-wise sentiment scores were calculated by linearly combining the keyword-weight vector and sentiment-score vector of each topic. The customer online review data in this paper was obtained from 2,719 reviews of TripAdvisor's Super 8 hotels, and after analysis, it was finally summarised into seven topics. The results of the analysis show that each topic has its own meaning and that the strengths and weaknesses of the hotel can be seen in each topic, thus providing professional assistance to the relevant hotel practitioners.

지구과학교육의 연구 동향: 데이터 마이닝 기법 잠재디리클레할당을 이용하여

곽민호(Kwak Minho),신윤주(Shin Yoonjoo),이진희(Lee Jinhee) 학습자중심교과교육학회 2019 학습자중심교과교육연구 Vol.19 No.18

본 연구는 지구과학교육의 연구 동향을 알아보기 위해 텍스트 마이닝 기법의 하나인 잠재디리클레할당(Latent Dirichlet Allocation: LDA)을 사용하여 2010년 1월부터 2019년 7월까지 약 10년 간 세 개의 지구과학교육 관련 저널에 게재된 연구 논문 들을 분석하였다. 잠재디리클레할당 방법은 주제와 단어 간의 관계와 각 문서에 나타나는 주제의 분포를 효과적으로 분석하는 텍스트 마이닝 방법의 하나로 다량의 문서를 처리하는 데에 널리 활용되고 있다. 연구 결과, 지난 10년 간 지구과학교육 연 구의 주된 주제는 학생 간 비교, STEAM, 학습 효과, 학습 자료, 학생의 개념 및 교사에 대한 연구인 것으로 나타났으며, 교과 내용으로는 지질학 및 천문학 연구가 많이 행해진 것으로 나타났다. 연구 대상으로는 초등학생 및 영재 연구가 활발하였다. 지난 10년 내에서 최근 들어 증가한 연구는 STEAM, 교사, 과학의 본성, 기존 연구분석, 스토리텔링 연구인 반면, 창의력에 대한 직접적인 연구 빈도는 줄어든 것으로 나타났다. This study aims to understand the research trend in Earth Science education. For this purpose, we analyzed research papers published in Earth Science education related journals from January 2010 to July 2019, utilizing an application of text mining, Latent Dirichlet Allocation (LDA). LDA, a popular method for analyzing massive documents, can effectively recognize relationships between topics and words and topic distribution in each document. The results suggested main topics of the research in Earth Science education as follows: comparisons between students, STEAM, learning effects, learning materials, concepts of students, and teachers. With regard to the contents of the curriculum, geology and astronomy are the main topics. In terms of study subjects, the research tends to be conducted on elementary school students and gifted students. Also, the most recent research trend showed that while studies including STEAM, teachers, nature of science, research review, and storytelling studies increase, studies on creativity decrease.

A comparison between logistic regression and neural networks in a constructed response item study

Minho Kwak,Chelwoo Park 한국데이터정보과학회 2019 한국데이터정보과학회지 Vol.30 No.5

The purpose of the study is to demonstrate the prediction quality of logistic regression and artificial neural networks. The main results of the study are the comparisons of the accuracy of both methods. The response variable of the model is a comment assignment by a human rater, and the four predictors are topic proportions estimated from latent Dirichlet allocation. The constructed models for both analyses are mainly concerned with predicting the comment assignment by using the topic proportions as the predictors. The results show that the accuracy of the test data set is generally higher than the accuracy of the cross-validation quality of the logistic regression, and these results are well matched with previous empirical studies. Also, although the use of this accuracy for practical purposes remains still questionable, the results reveal the potential utility the neural network if larger sample size is available in the future.

Analyzing genderless fashion trends of consumers’ perceptions on social media: using unstructured big data analysis through Latent Dirichlet Allocation‑based topic modeling

김효정,조인호,박민정 한국의류학회 2022 Fashion and Textiles Vol.9 No.1

After the development of Web 2.0 and social networks, analyzing consumers’ responses and opinions in real-time became profoundly important to gain business insights. This study aims to identify consumers’ preferences and perceptions of genderless fashion trends by text-mining, Latent Dirichlet Allocation-based topic modeling, and time-series linear regression analysis. Unstructured text data from consumer-posted sources, such as blogs and online communities, were collected from January 1, 2018 to December 31, 2020. We examined 9722 posts that included the keyword “genderless fashion” with Python 3.7 software. Results showed that consumers were interested in fragrances, fashion, and beauty brands and products. In particular, 18 topics were extracted: 13 were classifed as fashion categories and 5 were derived from beauty and fragrance sectors. Examining the genderless fashion trend development among consumers from 2018 to 2020, “perfume and scent” was revealed as the hot topic, whereas “bags,” “all-in-one skin care,” and “set-up suit” were cold topics, declining in popularity among consumers. The fndings contribute to contemporary fashion trends and provide in-depth knowledge about consumers’ perceptions using big data analysis methods and ofer insights into product development strategies.

A comparison between logistic regression and neural networks in a constructed response item study

곽민호,박철우 한국데이터정보과학회 2019 한국데이터정보과학회지 Vol.30 No.5

The purpose of the study is to demonstrate the prediction quality of logistic regression and artificial neural networks. The main results of the study are the comparisons of the accuracy of both methods. The response variable of the model is a comment assignment by a human rater, and the four predictors are topic proportions estimated from latent Dirichlet allocation. The constructed models for both analyses are mainly concerned with predicting the comment assignment by using the topic proportions as the predictors. The results show that the accuracy of the test data set is generally higher than the accuracy of the cross-validation quality of the logistic regression, and these results are well matched with previous empirical studies. Also, although the use of this accuracy for practical purposes remains still questionable, the results reveal the potential utility the neural network if larger sample size is available in the future.

Topic-wise sentiment analysis based on LASSO Regression and Latent Dirichlet Allocation: Focus on Customer reviews on Hotel

갈리총,정욱 한국서비스경영학회 2022 한국서비스경영학회 학술대회 Vol.2022 No.5

잠재 디리클레 할당 기반 토픽 모델링을 통한 건설재해 사례 분석

김창재 ( Kim Changjae ),김하림 ( Kim Harim ),이창수 ( Lee Changsu ),조훈희 ( Cho Hunhee ) 한국건축시공학회 2022 한국건축시공학회 학술발표대회 논문집 Vol.22 No.1

The construction industry has more safety accidents than other industries. Although there have been more attempts to reduce safety hazards in the industry such as the enforcement of the "Serious Accidents Punishment Act (SAPA)", construction accident has not been reduced enough. In this study, analysis of safety risk factors has been made through Latent Dirichlet Allocation (LDA)-based topic modeling. Risk analysis in construction site would be improved with natural language processing and topic modeling.

글로벌 화장품 브랜드의 소비자 만족도 분석: 텍스트마이닝 기반의 사용자 후기 분석을 중심으로

박재훈,김예림,강수빈 한국품질경영학회 2021 품질경영학회지 Vol.49 No.4

Purpose: This study introduces a systematic framework to evaluate service satisfaction of cosmetic brands through online review analysis utilizing Text-Mining technique. Methods: The framework assumes that the service satisfaction is evaluated by positive comments from online reviews. That is, the service satisfaction of a cosmetic brand is evaluated higher as more positive opinions are commented in the online reviews. This study focuses on two approaches. First, it collects online review comments from the top 50 global cosmetic brands and evaluates customer service satisfaction for each cosmetic brands by applying Sentimental Analysis and Latent Dirichlet Allocation. Second, it analyzes the determinants that induce or influence service satisfaction and suggests the guidelines for cosmetic brands with low satisfaction to improve their service satisfaction. Results: For the satisfaction evaluation, online review data were extracted from the top 50 global cosmetic brands in the world based on 2018 sales announced by Brand Finance in the UK. As a result of the satisfaction analysis, it was found that overall there were more positive opinions than negative opinions and the averages for polarity, subjectivity, positive ratio, and negative ratio were calculated as 0.50, 0.76, 0.57, and 0.19, respectively. Polarity, subjectivity and positive ratio showed the opposite pattern to negative ratio, and although there was a slight difference in fluctuation range and ranking between them, the patterns are almost same. Conclusion: The usefulness of the proposed framework was verified through case study. Although some studies have suggested a method to analyze online reviews, they didn't deal with the satisfaction evaluation among competitors and cause analysis. This study is different from previous studies in that it evaluates service satisfaction from a relative point of view among cosmetic brands and analyze determinants.

잠재디리클레할당 분석을 이용한 ‘노인일자리’ 관련 신문기사 토픽분석

이소정 한국디지털정책학회 2020 디지털융복합연구 Vol.18 No.10

This study aims to find the structure of social disussion on government ‘Senior job program’ by analyzing 1107 newspaper articles on ‘senior job program’ from 11 major newspaper articles and 8 financial newspapers. Topic modeling via latent dirichlet allocation model was employed for analysis and as result, 5 latent topics were extracted as follows : general information, local government project propaganda, senior life related issues, employment creation effect and market relations. Until 2015, most of the articles focused on the first two topics, indicating not much discourse was formed concerning the characteristics of the program. However, after 2015, the third topic started to increase and after the launch of Moon Jae In government, there has been a drastic increase in the employment creation related topic indicating that current social discourse mirrored by the media is definitely focused on employment creation aspect of senior job program. Based on the result, this study suggests the necessity to increase the quality and also enhance employment aspects of Senior job program. 본 연구는 노인일자리사업의 사회적 논의구조를 분석하기 위해 대표적인 대중매체인 신문기사에서 다루어지는 노인일자리 관련 주요 토픽들과 시계열적 특성을 분석하였다. 이를 위해 뉴스 통합 데이터베이스인 빅카인즈에 수록된 11개 중앙지와 8개 경제지의 노인일자리사업 관련 기사 1107개에 대해 잠재디리클레할당 방법을 이용한 토픽분석을 실시해 언론 기사에 내재된 노인일자리사업의 잠재토픽을 추출하였다. 분석결과 노인일자리사업에 대한 일반적 정보전달, 지자체 사업 홍보, 노후생활, 고용효과, 시장연계 등 5개의 잠재토픽이 추출되었는데 2015년까지 대부분의 언론 기사가 일반적 정보전달과 지자체 사업홍보에 국한되어 있어 노인일자리사업의 정체성에 대한 사회적 논의가 형성되지 못하였음을 알 수 있었던 반면 2015년 이후부터 노인일자리사업의 소득, 안전 등 노후생활 효과 관련 주제가 다루어지는 비중이 증가했으며 특히 문재인 정부 출범이후 고용효과와 관련된 기사가 압도적인 비중을 차지하게 되었음을 발견할 수 있었다. 본 연구는 이러한 결과에 근거해 향후 노인일자리사업의 질적측면 및 고용효과 측면을 증진시킬 수 있는 방안에 대한 고민의 필요성과 고용프레임 이외의 대안적 프레임 제시의 필요성을 제안하였다.