Abstract
Predicting stock price movement direction is a challenging task for financial investment. Previous researches focused on investigating the impacts of external factors (e.g., big events, economic influence and sentiments) in combination with the historical price to predict short-term stock price movement, while few researches leveraged the power of various relationships among enterprises. To bridge this gap, this research proposes power vector model and influence propagation model to mine the rich information in constructed Enterprise Knowledge Graph (EKG) for price movement prediction. In addition, Deep Neural Network (DNN) is introduced to train the model. The proposed model shows good prediction performance on the dataset of China top 500 enterprises.
Recommended Citation
Zhang, Yuxin; Yang, Kunlin; Du, Wei; and Xu, Wei, "Predicting Stock Price Movement Direction with Enterprise Knowledge Graph" (2018). PACIS 2018 Proceedings. 237.
https://aisel.aisnet.org/pacis2018/237