인문학
사회과학
자연과학
공학
의약학
농수해양학
예술체육학
복합학
지원사업
학술연구/단체지원/교육 등 연구자 활동을 지속하도록 DBpia가 지원하고 있어요.
커뮤니티
연구자들이 자신의 연구와 전문성을 널리 알리고, 새로운 협력의 기회를 만들 수 있는 네트워킹 공간이에요.
초록·키워드
Although stock price forecasting has been a traditional topic in the research domain of investment decision making, there have been many difficulties in forecasting stock price due to the unexpected rapid changes in stock prices. Recently, many researchers attempted to analyze sentiment in SNS data or news data to forecast stock price, but these researches have limitations that they used only one of sentiment data or KOSPI (Korea Composite Stock Price index) data in forecasting stock price. The aim of this paper is to propose new domain-specific sentiment dictionaries on stock price by using sentiment analysis, and acquire daily sentiment indices by analyzing the sentiment of news articles, and then use both of the sentiment data and KOSPI data together as input for data mining model for daily KOSPI forecasting, and finally improve the accuracy of forecasting the direction of KOSPI. TF-IDF weight was considered in building sentiment dictionaries and calculating daily sentiment indices by using domain-specific sentiment dictionaries. Our empirical result showed that in particular, a K-NN model with KOSPI and the sentiment data calculated by using both TF-IDF weights-based sentiment dictionary and the weights of news article itself in each news article data, had the accuracy of 68% and outperformed any other models in validation data.
본문·목차
인공지능 문자 인식 모델을 통해 추출된 텍스트로, 일부 오타나 오류가 포함될 수 있으나 지속적으로 개선 중입니다.
오류를 발견하셨다면 해당 부분을 드래그한 후 ' 를 통해 신고해주세요.
오류를 발견하셨다면 해당 부분을 드래그한 후 ' 를 통해 신고해주세요.