Start Date

14-12-2012 12:00 AM

Description

This research-in-progress paper attempts to identify events that have an effect on future oil price changes based on the predictive power of news messages on the oil market. For this purpose, text mining algorithms are used to determine whether news messages can be regarded as oil-relevant. Examining some 45 million news messages over a period of 8 years, a decision support system is constructed for early warning purposes. The system uses an indicator metric that triggers an alarm for events that will have an impact on the future (short-term) oil price. Relating to historic oil price series, regression analyses formally attest the predictive power of online news messages and thus the potential of the early warning system.

Share

COinS
 
Dec 14th, 12:00 AM

Towards an Oil Crisis Early Warning System based on Absolute News Volume

This research-in-progress paper attempts to identify events that have an effect on future oil price changes based on the predictive power of news messages on the oil market. For this purpose, text mining algorithms are used to determine whether news messages can be regarded as oil-relevant. Examining some 45 million news messages over a period of 8 years, a decision support system is constructed for early warning purposes. The system uses an indicator metric that triggers an alarm for events that will have an impact on the future (short-term) oil price. Relating to historic oil price series, regression analyses formally attest the predictive power of online news messages and thus the potential of the early warning system.