http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
거주 후 평가(POE) 방법으로써 오피니언 마이닝(Opinion Mining)의 활용가능성에 관한 연구
강현빈(Kang, Hyun-Bin),강주희(Kang, Ju-Hee),류예나(Ryu, Yeah-Na),손지연(Son, Ji-Youn),윤별이(Youn, Byeol-Yi),이준성(Yi, June-Seong) 대한건축학회 2017 대한건축학회 학술발표대회 논문집 Vol.37 No.2
Korean is a complex subtle language. Many Korean users upload personal stories to the Internet, but Korean language in the field of natural language processing is not developed due to the relatively low number of users. In this paper, we investigated the possibility of using Opinion Mining as a Post-Occupancy Evaluation method. The focus of this study is on how we can translate customer - provided housing experiences into useful information in data analysis. We analyzed the articles posted in the online community using the Python package called ‘KoNLPy’. In this process, I felt the difficulty of Korean language processing. In Korean natural language processing packages, it is possible to break up into an eulogy unit, but it is not possible to grasp the meaning of the correct word or to classify it into a proper morpheme. The construction industry in Korea is very conservative and it will not be easy to apply Opinion Mining to the practical stage. However, if the vocabulary network related to the residence experience is established and the skill level of observing the tendency of the people through the social network analysis is increased, the utilization of the natural language processing in the construction industry is expected to be very high.
Word2Vec 기반 데이터 시각화를 활용한 건설 재해관련 공공데이터 텍스트 분석
강현빈(Kang, Hyun-Bin),이준성(Yi, June-Seong) 대한건축학회 2018 대한건축학회 학술발표대회 논문집 Vol.38 No.2
The construction industry is an industry made up of environmental, physical, and human complex elements such as medicine, law, and criminology. Like similar industries, many data are being written in text form. This paper embedded the words appering in the public data of the construction disaster case provided by the Korean Occupational Safety and Health Agency, visualized and analyzed them. The Word2Vec algorithm is used for vectoring text data, and the t-SNE methodology is used for visualization. The results show that words with the same attributes were formed together without careful preprocessing.