Is this project an undergraduate, graduate, or faculty project?
Graduate
Loading...
individual
What campus are you from?
Daytona Beach
Authors' Class Standing
Graduate Student
Lead Presenter's Name
Amina Issoufou Anaroua
Faculty Mentor Name
Dr. Hari Adhikari
Abstract
The primary objective is to measure the impact of US airlines’ stock performance following aviation related news announcements. Our data account for aviation news of airlines, airports, regulations, safety, accidents, manufacturers, MRO, incidents, aviation training, general aviation and others from Aviation Voice. The amount of such data and documents is expected to be enormous. We use a natural language processing, Latent Dirichlet Allocation (LDA) to investigate and search for patterns that can explain the movement of US airline stock. First, we mine the aviation related data through text mining and topic modeling. Second, we employ the LDA model approach to help us identify and capture the extent of certain topics mentioned in aviation voice news releases. Finally, we use Event Study Methodology (ESM), which is a common method used to investigate stock price reactions to news announcements. We apply this method to discover the significance of the relationship between the stock return and the associated event. Eventually, we succeed in extracting 10 topics. As hypothesized, the impact of those topics varies greatly from topic to topic. Some topics have no significant effect on US Airline stocks while others such as aviation fuel price and air travel demand have significant effect.
Did this research project receive funding support from the Office of Undergraduate Research.
Yes, Student Internal Grant
Analysis of US Airline Stocks performance using Latent Dirichlet Allocation (LDA)
The primary objective is to measure the impact of US airlines’ stock performance following aviation related news announcements. Our data account for aviation news of airlines, airports, regulations, safety, accidents, manufacturers, MRO, incidents, aviation training, general aviation and others from Aviation Voice. The amount of such data and documents is expected to be enormous. We use a natural language processing, Latent Dirichlet Allocation (LDA) to investigate and search for patterns that can explain the movement of US airline stock. First, we mine the aviation related data through text mining and topic modeling. Second, we employ the LDA model approach to help us identify and capture the extent of certain topics mentioned in aviation voice news releases. Finally, we use Event Study Methodology (ESM), which is a common method used to investigate stock price reactions to news announcements. We apply this method to discover the significance of the relationship between the stock return and the associated event. Eventually, we succeed in extracting 10 topics. As hypothesized, the impact of those topics varies greatly from topic to topic. Some topics have no significant effect on US Airline stocks while others such as aviation fuel price and air travel demand have significant effect.