What if machines could discover human behavioral patterns better than experts? Traditional behavioral modeling in economics depends on costly manual refinement by domain experts, severely limiting scalability and discovery potential. We introduce AutoBM, an automated behavioral modeling framework leveraging large language models (LLMs) to systematically generate, evaluate, and refine interpretable behavioral models directly from human behavior data. AutoBM represents candidate models as structured natural language specifications, explicitly defining symbolic terms along with their tunable parameters, interpretations, and design rationales. AutoBM leverages LLMs to automatically translate each language specification into executable code, optimize tunable parameters, and evaluate model performance. Utilizing LLM-guided search strategies, AutoBM iteratively recombines and improves models at the term level, closely mirroring human expert practices. Experiments conducted across three distinct strategic environments (the ultimatum game, repeated rock-paper-scissors, and continuous double auctions) demonstrate that AutoBM-generated models consistently outperform leading manually crafted models, achieving significant improvements in prediction accuracy while maintaining clear interpretability. Our results demonstrate that automated frameworks can not only match but systematically exceed human expertise in behavioral modeling, fundamentally changing how we understand strategic human behavior.

An increasing number of mobile applications (abbrev. apps), like Pokemon Go and Snapchat, reward the users who physically visit some locations tagged as POIs (places-of-interest) by the apps. We study the novel POI-based collaboration between apps and venues (e.g., restaurants). On the one hand, an app charges a venue and tags the venue as a POI. The POI tag motivates users to visit the venue, which potentially increases the venue's sales. On the other hand, the venue can invest in the app-related infrastructure, which enables more users to use the app and further benefits the app's business. The apps' existing POI tariffs cannot fully incentivize the venue's infrastructure investment, and hence cannot lead to the most effective app-venue collaboration. We design an optimal two-part tariff, which charges the venue for becoming a POI, and subsidizes the venue every time a user interacts with the POI. The subsidy design efficiently incentivizes the venue's infrastructure investment, and we prove that our tariff achieves the highest app's revenue among a general class of tariffs. Furthermore, we derive some counter-intuitive guidelines for the POI-based collaboration. For example, a bandwidth-consuming app should collaborate with a low-quality venue (users have low utilities when consuming the venue's products).

Mobile Network Operators (MNOs) are providing more flexible wireless data services to attract subscribers and increase revenues. For example, the data trading market enables user-flexibility by allowing users to sell leftover data to or buy extra data from each other. The rollover mechanism enables time-flexibility by allowing a user to utilize his own leftover data from the previous month in the current month. In this paper, we investigate the economic viability of offering the data trading market together with the rollover mechanism, to gain a deeper understanding of the interrelationship between the user-flexibility and the time-flexibility. We formulate the interactions between the MNO and mobile users as a multi-slot dynamic game. Specifically, in each time slot (e.g., every day), the MNO first determines the selling and buying prices with the goal of revenue maximization, then each user decides his trading action (by solving a dynamic programming problem) to maximize his long-term payoff. Due to the availability of monthly data rollover, a user's daily trading decision corresponds to a dynamic programming problem with two time scales (i.e., day-to-day and month-to-month). Our analysis reveals an optimal trading policy with a target interval structure, specified by a buy-up-to threshold and a sell-down-to threshold in each time slot. Moreover, we show that the rollover mechanism makes users sell less and buy more data given the same trading prices, hence it increases the total demand while decreasing the total supply in the data trading market. Finally, numerical results based on real-world data unveil that the time-flexible rollover mechanism plays a positive role in the user-flexible data trading market, increasing the MNO's revenue by 25% and all users' payoff by 17% on average.

Many mobile applications (abbrev. apps) reward the users who physically visit some locations tagged as POIs (places-of-interest) by the apps. In this paper, we study the POI-based collaboration between apps and venues (e.g., restaurants and cafes). On the one hand, an app charges a venue and tags the venue as a POI, which attracts users to visit the venue and potentially increases the venue's sales. On the other hand, the venue can invest in the app-related infrastructure (e.g., Wi-Fi networks and smartphone chargers), which enhances the users' experience of using the app. However, the existing POI pricing schemes of the apps (e.g., Pokemon Go and Snapchat) cannot incentivize the venue's infrastructure investment, and hence cannot achieve the most effective app-venue collaboration. We model the interactions among an app, a venue, and users by a three-stage Stackelberg game, and design an optimal two-part pricing scheme for the app. This scheme has a charge-with-subsidy structure: the app first charges the venue for becoming a POI, and then subsidizes the venue every time a user interacts with the POI. Compared with the existing pricing schemes, our two-part pricing better incentivizes the venue's investment, attracts more users to interact with the POI, and achieves a much larger app revenue. We analyze the impacts of the app's and venue's characteristics on the app's optimal revenrevenueue, and show that the apps with small and large congestion effects should collaborate with opposite types of venues.

For Internet-based services, users' quality of service (QoS) depends on not only the available resource (capacity) but also the number of users who use the resource simultaneously (e.g., congestion effect). When a new Internet-based service provider first enters the market, there can be uncertainties regarding both the capacity and congestion, and hence the uncertainty of QoS. In this paper, we consider a participation-dependent social learning over the QoS through users' online reviews, where the QoS changes with the number of review participants. We study how such a learning process affects the provider's dynamic pricing strategy. With a simple two-period model, we analyze the strategic interactions between the provider and the users, and characterize the provider's optimal two-period dynamic pricing policy. Our results show that when the capacity is small or the users' prior QoS belief is high, the provider will choose a higher introductory price in the first period (than the price in the second period). This is in sharp contrast with the common practice of setting a lower introductory price to attract users (when congestion is not an issue). Furthermore, the learning process is beneficial to the provider with a large capacity.

This paper analyzes two pricing schemes commonly used in Wi-Fi markets: the flat-rate and the usage-based pricing. The flat-rate pricing encourages the maximum usage, while the usage-based pricing can flexibly attract more users especially those with low valuations in mobile Internet access. First, we use theoretical analysis to compare the two schemes and show that for a single provider in a market, as long as the Wi-Fi capacity is abundant, the flat-rate pricing leads to more revenue. Second, we study how a global provider (e.g., Skype) collaborates with this monopolist in each local market to provide a global Wi-Fi service. We formulate the interactions between the global and local providers as a dynamic game. In Stage I, the global provider bargains with the local provider in each market to determine the global Wi-Fi service price and revenue sharing agreement. In Stage II, local users and travelers choose local or global Wi-Fi services. We analytically show that the global provider prefers to use the usage-based pricing to avoid a severe competition with the local provider. At the equilibrium, the global provider always shares the majority of his revenue with the local provider to incentivize the cooperation. Finally, we analytically study how the interaction changes if the local market has more than one local provider. In this case, the global provider can integrate the coverages of multiple local providers and provide a better service. Compared to the local monopoly case, local market competition enables the global provider to share less revenue with each of the local providers. However, we numerically show that the global provider's revenue could decrease, as he shares his revenue with more providers and can only charge a lower price.

Train rescheduling is an important decision process in railway management. It aims to minimize the negative effects arising from the disturbances via real-time traffic management. Two main challenges are how to formulate the dynamic and complex rescheduling problem as an optimization model, and how to obtain a good solution within a short time limit. Focusing on the stochastic capacity recovery times of blocked tracks, we propose a new track-backup rescheduling (TBR) approach which optimally assigns each affected train a backup track, based on the estimation of recovery time, the original timetable, and track changing cost. Then, we formulate a mixed integer programming (MIP) model to obtain a conflict-free timetable which minimizes the delay cost and the expected track changing cost. A greedy algorithm is designed to reorder trains and reschedule the arrival and departure times, and then we use an MIP algorithm to solve the optimal track backup strategy. Based on the Beijing-Shanghai high-speed railway line, we conduct extensive experimental studies which show that the TBR approach can reduce the rescheduling cost by an average of 10.17% compared with traditional approaches. More important, the greedy-based algorithm is shown to be able to obtain good solutions (with an average error of only 2.85%) within 1.5 s, which implies the high potential of our approach in a real-time traffic management system where fast response is critical.

This paper analyzes two pricing schemes commonly used in WiFi markets: flat-rate pricing and usage-based pricing. The flat-free pricing encourages users to achieve the maximum WiFi usage and targets at users with high valuations in mobile Internet access, whereas the usage-based pricing is flexible to attract more users - even those with low valuations. First, we show that for a local provider, the flat-rate pricing provides more revenue than the usage-based pricing, which is consistent with the common practice in today's local markets. Second, we study how Skype may work with many local WiFi providers to provide a global WiFi service. We formulate the interactions between Skype, local providers, and users as a two-stage dynamic game. In Stage I, Skype bargains with each local provider to determine the global Skype WiFi service price and revenue sharing agreement; in Stage II, local users and travelers decide whether and how to use local or Skype WiFi service. Our analysis discovers two key insights behind Skype's current choice of usage-based pricing for its global WiFi service: to avoid severe competition with local providers and attract travelers to the service. We further show that at the equilibrium, Skype needs to share the majority of his revenue with a local provider to compensate the local provider's revenue loss due to competition. When there are more travelers or fewer local users, the competition between Skype and a local provider becomes less severe, and Skype can give away less revenue and reduce its usage-based price to attract more users.

Providing proper economic incentives to all parties involved is essential for the success of dynamic spectrum access. Market-driven secondary spectrum trading is an effective way to achieve this goal, where primary spectrum owners (POs) temporarily lease their licensed spectrum bands to unlicensed secondary users (SUs). In this paper, we consider the short-term secondary spectrum trading between one PO (seller) and multiple SUs (buyers) in a hybrid spectrum market with both guaranteed contracts (futures market) and spot transactions (spot market). In particular, we focus on the PO's expected profit maximization under stochastic network information. The optimal solution consists of (i) a policy that maximizes the ex-ante expected profit based on the stochastic distribution of network information, and (ii) a selling mechanism that determines the real-time allocation and charging based on the realized network information and the derived policy. We study the optimal solution systematically under both information symmetry and asymmetry, depending on whether the PO can observe the SUs' realized private information. Under information symmetry, we show that the optimal solution can be achieved by a perfect price discrimination mechanism, which maximizes both the PO's expected profit (optimality) and the social welfare (efficiency). Under information asymmetry, we propose an integrated contract and auction design-ContrAuction-to elicit SUs' private information effectively. We derive analytically the optimal ContrAuction mechanisms that maximize the PO's expected profit with and without the constraint of efficiency, and characterize systematically the tradeoff between the PO's profit and the social welfare.

Femtocells can effectively resolve the poor connectivity issue of indoor cellular users. This paper investigates the economic incentive for a cellular operator to add femtocell service on top of its existing macrocell service. We model the interactions between a cellular operator and users as a Stackelberg game: The operator first determines spectrum allocations and pricings of femtocell and macrocell services, and then heterogeneous users choose between the two services and the amount of resource to request. In the ideal case where the femtocell service has the same full spatial coverage as the macrocell service, we show that the operator will choose to provide femtocell service only, as this leads to a better user quality of service and a higher operator profit. However, if we impose the constraint that no users' payoffs decrease after introducing the femtocell service, then the operator will always continue providing the macrocell service (with or without the femtocell service). Furthermore, we study the impact of operational cost, limited coverage, and spatial reuse on femtocell service provision. As the operational cost increases, fewer users are served by femtocell service and the operator's profit decreases. When the femtocell service has limited spatial coverage, the operator always provides the macrocell service beside the femtocell service. However, when the coverage is high or the total resource is low, the operator will set the prices such that all users who can access femtocell will choose to use the femtocell service only. Finally, spatial reuse of spectrum will increase the efficiency of femtocell services and gives the operator more incentives to allocate spectrum to femtocells.

This paper presents a comprehensive analytical study of two competitive secondary operators' investment (i.e., spectrum leasing) and pricing strategies, taking into account operators' heterogeneity in leasing costs and users' heterogeneity in transmission power and channel conditions. We model the interactions between operators and users as a three-stage dynamic game, where operators simultaneously make spectrum leasing decisions in Stage I, and pricing decisions in Stage II, and then users make purchase decisions in Stage III. Using backward induction, we are able to completely characterize the dynamic game's equilibria. We show that both operators' investment and pricing equilibrium decisions process interesting threshold properties. For example, when the two operators' leasing costs are close, both operators will lease positive spectrum. Otherwise, one operator will choose not to lease and the other operator becomes the monopolist. For pricing, a positive pure strategy equilibrium exists only when the total spectrum investment of both operators is less than a threshold. Moreover, two operators always choose the same equilibrium price despite their heterogeneity in leasing costs. Each user fairly achieves the same service quality in terms of signal-to-noise ratio (SNR) at the equilibrium, and the obtained predictable payoff is linear in its transmission power and channel gain. We also compare the duopoly equilibrium with the coordinated case where two operators cooperate to maximize their total profit. We show that the maximum loss of total profit due to operators' competition is no larger than 25 percent. The users, however, always benefit from operators' competition in terms of their payoffs. We show that most of these insights are robust in the general SNR regime.

This paper studies the optimal investment and pricing decisions of a cognitive mobile virtual network operator (C-MVNO) under spectrum supply uncertainty. Compared with a traditional MVNO who often leases spectrum via long-term contracts, a C-MVNO can acquire spectrum dynamically in short-term by both sensing the empty “spectrum holes” of licensed bands and dynamically leasing from the spectrum owner. As a result, a C-MVNO can make flexible investment and pricing decisions to match demands of the secondary unlicensed users. Compared to dynamic spectrum leasing, spectrum sensing is typically cheaper, but the obtained useful spectrum amount is random due to primary licensed users' stochastic traffic. The C-MVNO needs to determine the optimal amounts of spectrum sensing and leasing by evaluating the trade-off between cost and uncertainty. The C-MVNO also needs to determine the optimal price to sell the spectrum to the secondary unlicensed users, taking into account wireless heterogeneity of users such as different maximum transmission power levels and channel gains. We model and analyze the interactions between the C-MVNO and secondary unlicensed users as a Stackelberg game. We show several interesting properties of the network equilibrium, including threshold structures of the optimal investment and pricing decisions, the independence of the optimal price on users' wireless characteristics, and guaranteed fair and predictable QoS among users. We prove that these properties hold for general SNR regime and general continuous distributions of sensing uncertainty. We show that spectrum sensing can significantly improve the C-MVNO's expected profit and users' payoffs.

This paper presents the first analytical study of optimal investment and pricing decisions of a cognitive mobile virtual network operator (C-MVNO) under spectrum supply uncertainty. Compared with a traditional MVNO who only obtains spectrum by long-term leasing contracts, a C-MVNO can acquire short-term spectrum by both sensing the empty "spectrum holes" of licensed bands and dynamically leasing from the spectrum owner. As a result, a C-MVNO can make flexible investment and pricing decisions to match the current demands of the secondary unlicensed users. Spectrum sensing is typically cheaper than dynamic spectrum leasing, but the obtained useful spectrum amount is random due to primary licensed users' stochastic traffic. The CMVNO needs to determine the optimal amounts of sensing and leasing spectrum, considering the trade-offs between cost and uncertainty. The C-MVNO also needs to determine the optimal retail price to sell the spectrum to the secondary unlicensed users, taking into account wireless heterogeneity of users such as different maximum transmission power levels and channel gains. We model and analyze these decisions and the interactions between the C-MVNO and secondary users as a multi-stage Stackelberg game. We show several interesting properties of the network equilibrium, such as threshold structures of the optimal investment and pricing decisions, independence between the optimal price and users' wireless characteristics, and fair and predictable spectrum allocations to the users. Compared with the traditional MVNO, spectrum sensing can significantly improve the C-MVNO's expected profit and users' payoffs.

Pinpoint is a promising first step towards using a rich model of task context in proactive and dynamic IR systems. Pinpoint allows a user to navigate decision tree representations of problem spaces, built by domain experts, while dynamically entering annotations specific to their problem. The system then automatically generates queries to information repositories based on both the user's annotations and location in the problem space, producing results that are both task focused and problem specific. Initial feedback from users and domain experts has been positive.