Thông tin chung
  English
  Đề tài NC khoa học
  Bài báo, báo cáo khoa học
  Hướng dẫn Sau đại học
  Sách và giáo trình
  Các học phần và môn giảng dạy
  Giải thưởng khoa học, Phát minh, sáng chế
  Khen thưởng
  Thông tin khác
  Tài liệu tham khảo
  Hiệu chỉnh
 
Số người truy cập: 106,957,439

 Market-Aware Sentiment Analysis for Stock Microblogs
Tác giả hoặc Nhóm tác giả: Đặng Trung Thành; Yongqiang Cheng; Ken Hawick
Nơi đăng: 26th IEEE International Conference on Automation and Computing (ICAC), United Kingdom; Số: 26;Từ->đến trang: 338-343;Năm: 2021
Lĩnh vực: Công nghệ thông tin; Loại: Bài báo khoa học; Thể loại: Quốc tế
TÓM TẮT
In financial markets, public sentiment acquired from microblogs allows understanding of traders’ attitudes, and hence, it can be utilized in market analysis and prediction. Current works on sentiment analysis for financial microblogs only focus on the microblogging messages themselves but tend to ignore their corresponding securities exchange data in the finance market when the messages were created. This study proposes an approach to utilize the contextual information extracted from the stock market data to improve sentiment classification performance for stock-related microblogging messages. Specifically, pre-trained LSTM encoders are employed to interpret and transform the end-of-day and intraday stock data into vectors which are then incorporated into the sentiment prediction model. A 3-step training strategy is proposed to improve the convergence and accuracy of the multi-input models. Results from experiments indicate that contextual information from the stock market data improves the prediction accuracy of the sentiment classification by about 2.7%, attributed to both the end-of-day and intraday stock market data.
ABSTRACT
In financial markets, public sentiment acquired from microblogs allows understanding of traders’ attitudes, and hence, it can be utilized in market analysis and prediction. Current works on sentiment analysis for financial microblogs only focus on the microblogging messages themselves but tend to ignore their corresponding securities exchange data in the finance market when the messages were created. This study proposes an approach to utilize the contextual information extracted from the stock market data to improve sentiment classification performance for stock-related microblogging messages. Specifically, pre-trained LSTM encoders are employed to interpret and transform the end-of-day and intraday stock data into vectors which are then incorporated into the sentiment prediction model. A 3-step training strategy is proposed to improve the convergence and accuracy of the multi-input models. Results from experiments indicate that contextual information from the stock market data improves the prediction accuracy of the sentiment classification by about 2.7%, attributed to both the end-of-day and intraday stock market data.
[ 2023\2023m02d02_13_25_392021123838.pdf ]
© Đại học Đà Nẵng
 
 
Địa chỉ: 41 Lê Duẩn Thành phố Đà Nẵng
Điện thoại: (84) 0236 3822 041 ; Email: dhdn@ac.udn.vn