Behavioural advertising in the public transit network. (September 2019)

• Record Type: Journal Article
• Authors: Faroqi, Hamed; Mesbah, Mahmoud; Kim, Jiwon
• Is Part Of: Research in transportation business & management. Volume 32(2019)
• Publication Date: 2019-09
• Subjects: Smart card data -- Behavioural advertising -- Optimization -- Travel behaviour -- Intelligent transportation -- Data mining; Transportation -- Research -- Periodicals; Transportation -- Management -- Periodicals; Transportation -- Management; Transportation -- Research; Periodicals; 388.068
• DOI: 10.1016/j.rtbm.2019.100421
• Languages: English
• ISSNs: 2210-5395

Abstract

Abstract: Public transit networks play a significant role for urban advertisers because a considerable number of residents in urban areas use public transport for their transportation needs. Automated Fare Collection (AFC) systems provide advertisers with valuable records (smart card data) of the boarding and alighting transactions of passengers in the public transit network. While the demographic attributes of passengers are missed by most AFC systems around the world, the systems can still help to reconstruct the activities and trips of passengers. The availability of smart card data has provided a unique opportunity to create detailed models of passengers' travel behaviour. Hence, it is now possible to develop behavioural advertising techniques in the public transit network based on ongoing activities of passengers. Behavioural advertising in the public transit network considers not only location and time of trips, but also duration and type of passengers' activities. This paper proposes and compares two behavioural advertising models based on the smart card data attributes. The first model, a trip-based one, targets trips of passengers, which means it indicates the maximum number of trips on which an advertisement should be viewed, according to the purpose of each trip. The second model, a passenger-based one, targets passengers by maximizing the number of passengers who will view an advertisement relevant to their trip. Both models are formulated as linear programming models. Both models are run on a case study basis to explicitly present the outcomes of the models and the differences between them. Outcomes of each model determine a set of prime time and locations for advertisements in the public transit network. While the trip-based model targets more relevant trips with simpler computations, the passenger-based model displays advertisements to a greater number of passengers, with more complex computations.