Improving Predictive Data Analytics inline with Airline Industry

Abstract

Air Transportation is one of the toughest and most dynamic industries in the world. The airline industryis the one where a few high value customers are more significant than many low value ones. Naturally, thewinners are those who can do predictive analytics on the data and provide a personalized flying experience.There exists a tough competition among air travel companies especially on their fare pricing, Cargo price,passenger’s time, safety, security and other related travel issues. Travel companies are applying multitude oftechniques to retain customers to their carriers by using business intelligence, CRM, revenues management,financial analysis, and social media and so on. Big data analytics is the processing of complex data sets todiscover hidden patterns, get useful correlations, extract market trends, and uncover customer preferences,product or services ratings and guiding principles of business information. Big data analytics can benefitairlines for effective marketing of services, better revenue, better customer service, improved operationalefficiency, competitive advantages over rival organizations and other business benefits. Fare processinginvolves extracting and consolidating information from external sources such as Agent systems, Externalpricing systems, and internal pricing systems such as Revenue Accounting, Forecasting, Inventory andYield Management systems. This data itself will run into tens of terabytes. Newer data sources such asSocial Media conversations about pricing decisions & competitors, and un-structured data in enterpriselike customer service e-mail data, call center logs etc. will push the data volumes even further. Analyzingthe customer’s data from various systems including Loyalty, CRM, Sales & Marketing etc. and data frompartners systems along with data from social media based on customer social profile can help airlines tocreate deeper personalization preference of customers and also understand their current social status andpreferences. The research study proposed Big Data Architecture leveraging Apache Hadoop and Apache Sparkstack provides ability to extract, aggregate, load and process large volumes of data in a distributed mannerwhich in turn reduce the complexity and overall turn-around time in processing large volumes of data. Thebuilt predictive model predict the fare price based on the historical big data after finding high value customers.