Type recognition of financial events by incorporating text and time-series features

The investors in financial market have shown great concerns in the events that may cause fluctuations in the capital market. Traditional event detection and type recognition methods were majorly based on text processing techniques while few research considers the financial time-series features. As we know, there are large amount of financial time-series data available such as stock transaction data. These data may be regarded as the capital market response to specific financial events. By taking these factors into consideration, this paper presents a financial event recognition approach by incorporating text and time-series features. By following the event study method, the influences of specific type of event to stock index are estimated by analyzing the changes of average abnormal returns and average cumulative abnormal returns. In this way, the time-series features related to specific financial event types are obtained. These features are incorporated with text features in a support vector machine based classifier to recognize the types of financial events. The experiment results show that the performance of financial event type recognition is improved by incorporating text and time-series features.