Cuihong Li | University of Connecticut (original) (raw)
Papers by Cuihong Li
We provide an integrative solution for negotiations in the Navy detailing process considering the... more We provide an integrative solution for negotiations in the Navy detailing process considering the uncertain and dynamic outside options. The outside options influence the negotiation strategies via the impact on the reservation prices. The solution is composed of three modular models: single-threaded negotiations, synchronized multithreaded negotiations, and dynamic multi-threaded negotiations. The single-threaded negotiation model provides the negotiation strategy given the reservation price. The other two models calculate the reservation price of a negotiation thread based on the model of the outside options by viewing all outside options as a multi-threaded negotiation process. The model of synchronized multi-threaded negotiations considers the presence of concurrently available outside options and provides an approach to estimate the outcome when the threads are known. The model of dynamic multi-threaded negotiations expands the synchronized model by considering the uncertain outside options that may come dynamically in the future. We propose two effective negotiation strategies, the time-dependent strategy and Bayesian learning strategy, from the AI field. Four heuristic approaches are designed to estimate the expected utility from a synchronized multi-threaded negotiation. A Poisson process is used to model the random sequential arrivals, and formulas are provided to calculate the expected utility from a negotiation process when uncertain outside options may come in the future.
SSRN Electronic Journal, 2000
ABSTRACT We study a buyer's sourcing strategy along two dimensions: the supply base desig... more ABSTRACT We study a buyer's sourcing strategy along two dimensions: the supply base design and the pricing mechanism, considering supplier competition and cost-reduction effort. The supply base design concerns the number of suppliers (one or two) included in the supply base, and the capacity to be invested in each supplier. The pricing mechanism determines the timing of the price decisions, with the buyer making price commitments before suppliers exert cost-reduction efforts that may be renegotiated afterwards. We find that symmetric capacity investment in suppliers and low price commitments (more likely to be renegotiated) are effective at fostering supplier competition, while asymmetric investment and high price commitments (less likely to be renegotiated) are better at motivating supplier effort. A complementary relationship exists between the supply base design and pricing mechanism: A more symmetric supply base should be combined with lower price commitments leading to more renegotiation opportunities. This results in three possible sourcing structures: sole sourcing (investing capacity in a single supplier and forming price with ex ante commitments), symmetric dual sourcing (investing equal capacity in both suppliers and forming price with ex post negotiations), and asymmetric dual sourcing (investing positive but unequal capacities in two (ex ante identical) suppliers and forming price using both ex ante commitments and ex post (re)negotiations). We characterize the conditions for each structure, and identify a strategic role of capacity investment.
SSRN Electronic Journal, 2000
ABSTRACT Buyers can manage product quality sourced from suppliers in several ways: they can impro... more ABSTRACT Buyers can manage product quality sourced from suppliers in several ways: they can improve the incoming quality directly by cooperating with suppliers to improve a process or product, they can improve the incoming quality indirectly by incentivizing suppliers' quality-improvement efforts, or they can control the outgoing quality by inspecting incoming units. In this paper, we study the buyer's use of these three strategies -- cooperation, incentivization, and inspection -- to manage sourced quality. Specifically, we consider a general effect relationship between the buyer and supplier efforts, allowing for substitutable or complementary quality-improvement effects of these efforts. Our results establish a strategic relationship between these efforts, which may differ from their effect relationship due to the inclusion of endogenous inspection. We also identify the ways in which cooperation influences the incentive structure, which may drive cooperation positively or negatively depending on the effect relationship of efforts. Finally, we compare the results with observable and unobservable supplier efforts, characterizing the impact of supplier effort observability.
SSRN Electronic Journal, 2000
This paper studies the pre-order strategy that a seller may use to sell a perishable product in a... more This paper studies the pre-order strategy that a seller may use to sell a perishable product in an uncertain market with heterogeneous consumers. By accepting pre-orders, the seller is able to obtain advance demand information for inventory planning and price-discriminate the consumers. Given the pre-order option, the consumers react strategically by optimizing the timing of purchase. We find that accurate demand information may improve the availability of the product, which undermines the seller's ability to charge a high pre-order price. As a result, advance demand information may hurt the seller's profit due to its negative impact for the pre-order season. This cautions the seller about a potential conflict between the benefits of advance demand information and price discrimination when facing strategic consumers. A common practice to contain consumers' strategic waiting is to offer price guarantees that compensate pre-order consumers in case of a later price cut. Under price guarantees, the seller will reduce price in the regular season only if the pre-order demand is low; however, such advance information implies weak demand in the regular season as well. This means that the seller can no longer benefit from a high demand in the regular season. Therefore, under price guarantees, more accurate advance demand information may still hurt the seller's profit due to its adverse impact for the regular season. We also investigate the seller's strategy choice in such a setting (i.e., whether the pre-order option should be offered and whether it should be coupled with price guarantees) and find that the answer depends on the relative sizes of the heterogeneous consumer segments.
SSRN Electronic Journal, 2000
ABSTRACT Consider a buyer, facing uncertain demand, who sources from multiple suppliers via onlin... more ABSTRACT Consider a buyer, facing uncertain demand, who sources from multiple suppliers via online procurement auctions (open descending price-only auctions). The suppliers have heterogeneous production costs, which are private information, and the winning supplier has to invest in production capacity before the demand uncertainty is resolved. The buyer chooses to offer a push or pull contract, for which the single price and winning supplier are determined via the auction. We show that with a pull contract, the buyer does not necessarily benefit from a larger number of suppliers participating in the auction, due to the negative effect of supplier competition on the incentive of supplier capacity investment. We thus propose an enhanced pull mechanism that mitigates this effect with a floor price. We then analyze and compare the outcomes of auctions for push and (enhanced) pull contracts, establishing when one form is preferred over the other based on the buyer's profits. We also compare our simple, price-only push and pull contract auctions to the optimal mechanisms, benchmarking the performance of the simple mechanisms as well as establishing the relative importance of auction design and contract design in procurement auctions.
Production and Operations Management, 2011
ABSTRACT Consider a buyer, facing uncertain demand, who sources from multiple suppliers via onlin... more ABSTRACT Consider a buyer, facing uncertain demand, who sources from multiple suppliers via online procurement auctions (open descending price-only auctions). The suppliers have heterogeneous production costs, which are private information, and the winning supplier has to invest in production capacity before the demand uncertainty is resolved. The buyer chooses to offer a push or pull contract, for which the single price and winning supplier are determined via the auction. We show that with a pull contract, the buyer does not necessarily benefit from a larger number of suppliers participating in the auction, due to the negative effect of supplier competition on the incentive of supplier capacity investment. We thus propose an enhanced pull mechanism that mitigates this effect with a floor price. We then analyze and compare the outcomes of auctions for push and (enhanced) pull contracts, establishing when one form is preferred over the other based on the buyer's profits. We also compare our simple, price-only push and pull contract auctions to the optimal mechanisms, benchmarking the performance of the simple mechanisms as well as establishing the relative importance of auction design and contract design in procurement auctions.
Manufacturing & Service Operations Management, 2013
This paper studies the pre-order strategy that a seller may use to sell a perishable product in a... more This paper studies the pre-order strategy that a seller may use to sell a perishable product in an uncertain market with heterogeneous consumers. By accepting pre-orders, the seller is able to obtain advance demand information for inventory planning and price-discriminate the consumers. Given the pre-order option, the consumers react strategically by optimizing the timing of purchase. We find that accurate demand information may improve the availability of the product, which undermines the seller's ability to charge a high pre-order price. As a result, advance demand information may hurt the seller's profit due to its negative impact for the pre-order season. This cautions the seller about a potential conflict between the benefits of advance demand information and price discrimination when facing strategic consumers. A common practice to contain consumers' strategic waiting is to offer price guarantees that compensate pre-order consumers in case of a later price cut. Under price guarantees, the seller will reduce price in the regular season only if the pre-order demand is low; however, such advance information implies weak demand in the regular season as well. This means that the seller can no longer benefit from a high demand in the regular season. Therefore, under price guarantees, more accurate advance demand information may still hurt the seller's profit due to its adverse impact for the regular season. We also investigate the seller's strategy choice in such a setting (i.e., whether the pre-order option should be offered and whether it should be coupled with price guarantees) and find that the answer depends on the relative sizes of the heterogeneous consumer segments.
Autonomous Agents & Multiagent Systems/International Conference on Autonomous Agents, 2004
Most existing research on peer-to-peer systems focuses on protocol design. In this paper, we cons... more Most existing research on peer-to-peer systems focuses on protocol design. In this paper, we consider the issue of free riding in peer-to-peer referral systems. Free riders are agents that refuse either to answer a query or to give referrals. Free riding is detrimental to the system, since it may prevent requesters from finding high quality providers efficiently. To mitigate the issue of free riding, we propose a dynamic pricing mechanism to motivate the agents to behave rationally. Service providers learn appropriate prices of referrals and answers in order to maximize their payoffs through stochastic iterative learning algorithms.
Hawaii International Conference on System Sciences, 2005
We present a model for bilateral negotiations that considers the uncertain and dynamic outside op... more We present a model for bilateral negotiations that considers the uncertain and dynamic outside options. Outside options affect the negotiation strategies via their impact on the reservation price. The model is composed of three modules, single-threaded negotiations, synchronized multi-threaded negotiations, and dynamic multi-threaded negotiations. These three models embody increased sophistication and complexity. The single-threaded negotiation model provides negotiation strategies without specifically considering outside options. The model of synchronized multi-threaded negotiations builds on the single-threaded negotiation model and considers the presence of concurrently existing outside options. The model of dynamic multi-threaded negotiations expands the synchronized multi-threaded model by considering the uncertain outside options thatm ay come dynamically in the future. Experimental analysis is provided to characterize the impact of outside options on the reservation price and thus on the negotiation strategy. The results show that the utility of a negotiator improves significantly if she considers outside options, and the average utility is higher when she both considers the concurrent outside options and foresees future options.
Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems, 2004
Most existing research on peer-to-peer systems focuses on protocol design. In this paper, we cons... more Most existing research on peer-to-peer systems focuses on protocol design. In this paper, we consider the issue of free riding in peer-to-peer referral systems. Free riders are agents that refuse either to answer a query or to give referrals. Free riding is detrimental to the system, since it may prevent requesters from finding high quality providers efficiently. To mitigate the issue of free riding, we propose a dynamic pricing mechanism to motivate the agents to behave rationally. Service providers learn appropriate prices of referrals and answers in order to maximize their payoffs through stochastic iterative learning algorithms.
Group Decision and Negotiation, 2006
Multi-attribute negotiation is an important mechanism for distributed decision makers to reach ag... more Multi-attribute negotiation is an important mechanism for distributed decision makers to reach agreements in real-world situations. It allows the possibility of reaching “win-win” solutions for both parties, who trade off different attributes in a solution. Existing research on multi-attribute negotiations has mainly focused on the situations when negotiation parties have complete information about each other's preference. This paper presents a model with incomplete information, while considering Pareto-efficiency and computational efficiency. A non-biased mediator, who applies query learning to maintain near Pareto-efficiency without heavy computation, is adopted in the model. In addition, the mediating mechanism proposed in the model overcomes the difficulty of preference elicitation which usually arises in the preliminary step of a multi-attribute negotiation. Our model also reduces the negotiation complexity by decomposing the original n-dimensional negotiation space into a sequence of negotiation base lines. Agents can negotiate upon a base line with rather simple strategies. The experimental results show that near Pareto-efficient agreements can be reached effectively.
Proceedings of the fifteenth annual ACM symposium on Parallel algorithms and architectures - SPAA '03, 2003
We consider the practical problem of task assignment in a server farm, where each arriving job is... more We consider the practical problem of task assignment in a server farm, where each arriving job is immediately dispatched to a server in the farm. We look at the benefit of cycle stealing at the point of the dispatcher, where jobs normally destined for one machine may be routed to a different machine if it is idle. The analysis uses a technique which we refer to as dimensionality reduction via busy period transitions. Our analysis is approximate, but can be made as close to exact as desired, and is validated via simulation. Results show that the beneficiaries of the idle cycles can benefit unboundedly, due to an increase in their stability region, while the donors are only slightly penalized. These results still hold even when there is only one donor server and 20 beneficiary servers stealing its idle cycles.
23rd International Conference on Distributed Computing Systems, 2003. Proceedings., 2003
... prove to be very difficult: Many of the long-standing open questions in queueing theory invol... more ... prove to be very difficult: Many of the long-standing open questions in queueing theory involve the performance analysis of task assignment policies. ... consistent with validated stochas-tic models used to study a high-volume Web sites [10, 20], studies of scalable systems for ...
Proceedings of the fifteenth annual ACM symposium on Parallel algorithms and architectures - SPAA '03, 2003
We consider the practical problem of task assignment in a server farm, where each arriving job is... more We consider the practical problem of task assignment in a server farm, where each arriving job is immediately dispatched to a server in the farm. We look at the benefit of cycle stealing at the point of the dispatcher, where jobs normally destined for one machine may be routed to a different machine if it is idle. The analysis uses a technique which we refer to as dimensionality reduction via busy period transitions. Our analysis is approximate, but can be made as close to exact as desired, and is validated via simulation. Results show that the beneficiaries of the idle cycles can benefit unboundedly, due to an increase in their stability region, while the donors are only slightly penalized. These results still hold even when there is only one donor server and 20 beneficiary servers stealing its idle cycles.
In an electronic marketplace coalition formation allows buyers to enjoy a price discount for each... more In an electronic marketplace coalition formation allows buyers to enjoy a price discount for each item while combinatorial auction enables buyers to place bids for a bundle of items that are complementary. Coalition formation and combinatorial auction both help to improve the e ciency of a market and have received much attention from economists and computer scientists. But neither in laboratories nor in practice has there been literature on the situations where both coalition formation and combinatorial auctions exist. In this paper we consider an e-market where each buyer places a bid on a combination of items with a reservation cost, and sellers o er price discounts for each item based on volumes. We call coalition formation under this condition a Combinatorial Coalition Formation CCF problem since coalition formation is motivated by price discounts on single items while multiple items are complementary for buyers. By arti cially dividing the reservation cost of each buyer appropriately among the items we can construct optimal coalitions with respect to each item. We then try to make these coalitions satisfy the complementarity of the items, and thus induce the optimal solution. Based on this idea we present polynomial-time algorithms to nd a semi-optimal solution of CCF and a payo division scheme that is in the core of the coalition when linear price functions are applied, and in the pseudo-core when general price functions are applied. Simulation results show that the algorithms obtain solutions in a satisfactory ratio to the optimal value.
Proceedings of the 38th Annual Hawaii International Conference on System Sciences, 2005
We present a model for bilateral negotiations that considers the uncertain and dynamic outside op... more We present a model for bilateral negotiations that considers the uncertain and dynamic outside options. Outside options affect the negotiation strategies via their impact on the reservation price. The model is composed of three modules, single-threaded negotiations, synchronized multi-threaded negotiations, and dynamic multi-threaded negotiations. These three models embody increased sophistication and complexity. The single-threaded negotiation model provides negotiation strategies without specifically considering outside options. The model of synchronized multi-threaded negotiations builds on the single-threaded negotiation model and considers the presence of concurrently existing outside options. The model of dynamic multi-threaded negotiations expands the synchronized multi-threaded model by considering the uncertain outside options that may come dynamically in the future. Experimental analysis is provided to characterize the impact of outside options on the reservation price and thus on the negotiation strategy. The results show that the utility of a negotiator improves significantly if she considers outside options, and the average utility is higher when she both considers the concurrent outside options and foresees future options.
Multiagent and Grid Systems, 2008
Most of the prior work on multi-attribute negotiations usually assumes that agents have linear ad... more Most of the prior work on multi-attribute negotiations usually assumes that agents have linear additive utility functions on the attributes (e.g. [9, 11, 13]), or focuses on issue-by-issuenegotiations [1, 3]. The work that addresses complex utility functions and Pareto optimality, either assumes complete information, cooperative agents, the presence of a non-biased mediator [7], or makes the assumption that attributes have binary values [6,16,24]. This chapter aims to propose a model that can help agents negotiate multiple attributes with complex utility functions and incomplete information, but still maintains Pareto optimality. Thus, this chapter contributes to the existing literature in the following aspects: first, focusing on decentralized negotiations, our model can be applied to the situations with self-interested agents; second, the model allows agents to negotiate multiple attributes simultaneously and considers Pareto optimality of the negotiation outcome; third, the model can be applied to incomplete information scenarios where each agent knows neither the utility function nor the negotiation strategy of her opponent; and finally, the model can be applied to general situations where agents have complex utility functions in continuous negotiation domains. The rest of the chapter is organized as follows. Section 2 reviews the related work. Section 3 presents the model. Experimental analysis is provided in Sect. 4, and in Sect. 5 we conclude.
Electronic Commerce Research and Applications, 2004
We study the mechanism design problem of coalition formation and cost sharing in a group-buying e... more We study the mechanism design problem of coalition formation and cost sharing in a group-buying electronic marketplace, where buyers can form coalitions to take advantage of volume based discounts. The desirable mechanism properties include stability (in the core), and incentive compatibility with good efficiency. We show the impossibility to simultaneously satisfy efficiency, budget balance and individual rationality at a Bayesian–Nash equilibrium, and propose a mechanism in the core of the game. We then present and evaluate a group of reasonable mechanisms. Empirical results show positive correlation between stability and incentive compatibility (which is in turn related to efficiency).
We provide an integrative solution for negotiations in the Navy detailing process considering the... more We provide an integrative solution for negotiations in the Navy detailing process considering the uncertain and dynamic outside options. The outside options influence the negotiation strategies via the impact on the reservation prices. The solution is composed of three modular models: single-threaded negotiations, synchronized multithreaded negotiations, and dynamic multi-threaded negotiations. The single-threaded negotiation model provides the negotiation strategy given the reservation price. The other two models calculate the reservation price of a negotiation thread based on the model of the outside options by viewing all outside options as a multi-threaded negotiation process. The model of synchronized multi-threaded negotiations considers the presence of concurrently available outside options and provides an approach to estimate the outcome when the threads are known. The model of dynamic multi-threaded negotiations expands the synchronized model by considering the uncertain outside options that may come dynamically in the future. We propose two effective negotiation strategies, the time-dependent strategy and Bayesian learning strategy, from the AI field. Four heuristic approaches are designed to estimate the expected utility from a synchronized multi-threaded negotiation. A Poisson process is used to model the random sequential arrivals, and formulas are provided to calculate the expected utility from a negotiation process when uncertain outside options may come in the future.
SSRN Electronic Journal, 2000
ABSTRACT We study a buyer's sourcing strategy along two dimensions: the supply base desig... more ABSTRACT We study a buyer's sourcing strategy along two dimensions: the supply base design and the pricing mechanism, considering supplier competition and cost-reduction effort. The supply base design concerns the number of suppliers (one or two) included in the supply base, and the capacity to be invested in each supplier. The pricing mechanism determines the timing of the price decisions, with the buyer making price commitments before suppliers exert cost-reduction efforts that may be renegotiated afterwards. We find that symmetric capacity investment in suppliers and low price commitments (more likely to be renegotiated) are effective at fostering supplier competition, while asymmetric investment and high price commitments (less likely to be renegotiated) are better at motivating supplier effort. A complementary relationship exists between the supply base design and pricing mechanism: A more symmetric supply base should be combined with lower price commitments leading to more renegotiation opportunities. This results in three possible sourcing structures: sole sourcing (investing capacity in a single supplier and forming price with ex ante commitments), symmetric dual sourcing (investing equal capacity in both suppliers and forming price with ex post negotiations), and asymmetric dual sourcing (investing positive but unequal capacities in two (ex ante identical) suppliers and forming price using both ex ante commitments and ex post (re)negotiations). We characterize the conditions for each structure, and identify a strategic role of capacity investment.
SSRN Electronic Journal, 2000
ABSTRACT Buyers can manage product quality sourced from suppliers in several ways: they can impro... more ABSTRACT Buyers can manage product quality sourced from suppliers in several ways: they can improve the incoming quality directly by cooperating with suppliers to improve a process or product, they can improve the incoming quality indirectly by incentivizing suppliers' quality-improvement efforts, or they can control the outgoing quality by inspecting incoming units. In this paper, we study the buyer's use of these three strategies -- cooperation, incentivization, and inspection -- to manage sourced quality. Specifically, we consider a general effect relationship between the buyer and supplier efforts, allowing for substitutable or complementary quality-improvement effects of these efforts. Our results establish a strategic relationship between these efforts, which may differ from their effect relationship due to the inclusion of endogenous inspection. We also identify the ways in which cooperation influences the incentive structure, which may drive cooperation positively or negatively depending on the effect relationship of efforts. Finally, we compare the results with observable and unobservable supplier efforts, characterizing the impact of supplier effort observability.
SSRN Electronic Journal, 2000
This paper studies the pre-order strategy that a seller may use to sell a perishable product in a... more This paper studies the pre-order strategy that a seller may use to sell a perishable product in an uncertain market with heterogeneous consumers. By accepting pre-orders, the seller is able to obtain advance demand information for inventory planning and price-discriminate the consumers. Given the pre-order option, the consumers react strategically by optimizing the timing of purchase. We find that accurate demand information may improve the availability of the product, which undermines the seller's ability to charge a high pre-order price. As a result, advance demand information may hurt the seller's profit due to its negative impact for the pre-order season. This cautions the seller about a potential conflict between the benefits of advance demand information and price discrimination when facing strategic consumers. A common practice to contain consumers' strategic waiting is to offer price guarantees that compensate pre-order consumers in case of a later price cut. Under price guarantees, the seller will reduce price in the regular season only if the pre-order demand is low; however, such advance information implies weak demand in the regular season as well. This means that the seller can no longer benefit from a high demand in the regular season. Therefore, under price guarantees, more accurate advance demand information may still hurt the seller's profit due to its adverse impact for the regular season. We also investigate the seller's strategy choice in such a setting (i.e., whether the pre-order option should be offered and whether it should be coupled with price guarantees) and find that the answer depends on the relative sizes of the heterogeneous consumer segments.
SSRN Electronic Journal, 2000
ABSTRACT Consider a buyer, facing uncertain demand, who sources from multiple suppliers via onlin... more ABSTRACT Consider a buyer, facing uncertain demand, who sources from multiple suppliers via online procurement auctions (open descending price-only auctions). The suppliers have heterogeneous production costs, which are private information, and the winning supplier has to invest in production capacity before the demand uncertainty is resolved. The buyer chooses to offer a push or pull contract, for which the single price and winning supplier are determined via the auction. We show that with a pull contract, the buyer does not necessarily benefit from a larger number of suppliers participating in the auction, due to the negative effect of supplier competition on the incentive of supplier capacity investment. We thus propose an enhanced pull mechanism that mitigates this effect with a floor price. We then analyze and compare the outcomes of auctions for push and (enhanced) pull contracts, establishing when one form is preferred over the other based on the buyer's profits. We also compare our simple, price-only push and pull contract auctions to the optimal mechanisms, benchmarking the performance of the simple mechanisms as well as establishing the relative importance of auction design and contract design in procurement auctions.
Production and Operations Management, 2011
ABSTRACT Consider a buyer, facing uncertain demand, who sources from multiple suppliers via onlin... more ABSTRACT Consider a buyer, facing uncertain demand, who sources from multiple suppliers via online procurement auctions (open descending price-only auctions). The suppliers have heterogeneous production costs, which are private information, and the winning supplier has to invest in production capacity before the demand uncertainty is resolved. The buyer chooses to offer a push or pull contract, for which the single price and winning supplier are determined via the auction. We show that with a pull contract, the buyer does not necessarily benefit from a larger number of suppliers participating in the auction, due to the negative effect of supplier competition on the incentive of supplier capacity investment. We thus propose an enhanced pull mechanism that mitigates this effect with a floor price. We then analyze and compare the outcomes of auctions for push and (enhanced) pull contracts, establishing when one form is preferred over the other based on the buyer's profits. We also compare our simple, price-only push and pull contract auctions to the optimal mechanisms, benchmarking the performance of the simple mechanisms as well as establishing the relative importance of auction design and contract design in procurement auctions.
Manufacturing & Service Operations Management, 2013
This paper studies the pre-order strategy that a seller may use to sell a perishable product in a... more This paper studies the pre-order strategy that a seller may use to sell a perishable product in an uncertain market with heterogeneous consumers. By accepting pre-orders, the seller is able to obtain advance demand information for inventory planning and price-discriminate the consumers. Given the pre-order option, the consumers react strategically by optimizing the timing of purchase. We find that accurate demand information may improve the availability of the product, which undermines the seller's ability to charge a high pre-order price. As a result, advance demand information may hurt the seller's profit due to its negative impact for the pre-order season. This cautions the seller about a potential conflict between the benefits of advance demand information and price discrimination when facing strategic consumers. A common practice to contain consumers' strategic waiting is to offer price guarantees that compensate pre-order consumers in case of a later price cut. Under price guarantees, the seller will reduce price in the regular season only if the pre-order demand is low; however, such advance information implies weak demand in the regular season as well. This means that the seller can no longer benefit from a high demand in the regular season. Therefore, under price guarantees, more accurate advance demand information may still hurt the seller's profit due to its adverse impact for the regular season. We also investigate the seller's strategy choice in such a setting (i.e., whether the pre-order option should be offered and whether it should be coupled with price guarantees) and find that the answer depends on the relative sizes of the heterogeneous consumer segments.
Autonomous Agents & Multiagent Systems/International Conference on Autonomous Agents, 2004
Most existing research on peer-to-peer systems focuses on protocol design. In this paper, we cons... more Most existing research on peer-to-peer systems focuses on protocol design. In this paper, we consider the issue of free riding in peer-to-peer referral systems. Free riders are agents that refuse either to answer a query or to give referrals. Free riding is detrimental to the system, since it may prevent requesters from finding high quality providers efficiently. To mitigate the issue of free riding, we propose a dynamic pricing mechanism to motivate the agents to behave rationally. Service providers learn appropriate prices of referrals and answers in order to maximize their payoffs through stochastic iterative learning algorithms.
Hawaii International Conference on System Sciences, 2005
We present a model for bilateral negotiations that considers the uncertain and dynamic outside op... more We present a model for bilateral negotiations that considers the uncertain and dynamic outside options. Outside options affect the negotiation strategies via their impact on the reservation price. The model is composed of three modules, single-threaded negotiations, synchronized multi-threaded negotiations, and dynamic multi-threaded negotiations. These three models embody increased sophistication and complexity. The single-threaded negotiation model provides negotiation strategies without specifically considering outside options. The model of synchronized multi-threaded negotiations builds on the single-threaded negotiation model and considers the presence of concurrently existing outside options. The model of dynamic multi-threaded negotiations expands the synchronized multi-threaded model by considering the uncertain outside options thatm ay come dynamically in the future. Experimental analysis is provided to characterize the impact of outside options on the reservation price and thus on the negotiation strategy. The results show that the utility of a negotiator improves significantly if she considers outside options, and the average utility is higher when she both considers the concurrent outside options and foresees future options.
Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems, 2004
Most existing research on peer-to-peer systems focuses on protocol design. In this paper, we cons... more Most existing research on peer-to-peer systems focuses on protocol design. In this paper, we consider the issue of free riding in peer-to-peer referral systems. Free riders are agents that refuse either to answer a query or to give referrals. Free riding is detrimental to the system, since it may prevent requesters from finding high quality providers efficiently. To mitigate the issue of free riding, we propose a dynamic pricing mechanism to motivate the agents to behave rationally. Service providers learn appropriate prices of referrals and answers in order to maximize their payoffs through stochastic iterative learning algorithms.
Group Decision and Negotiation, 2006
Multi-attribute negotiation is an important mechanism for distributed decision makers to reach ag... more Multi-attribute negotiation is an important mechanism for distributed decision makers to reach agreements in real-world situations. It allows the possibility of reaching “win-win” solutions for both parties, who trade off different attributes in a solution. Existing research on multi-attribute negotiations has mainly focused on the situations when negotiation parties have complete information about each other's preference. This paper presents a model with incomplete information, while considering Pareto-efficiency and computational efficiency. A non-biased mediator, who applies query learning to maintain near Pareto-efficiency without heavy computation, is adopted in the model. In addition, the mediating mechanism proposed in the model overcomes the difficulty of preference elicitation which usually arises in the preliminary step of a multi-attribute negotiation. Our model also reduces the negotiation complexity by decomposing the original n-dimensional negotiation space into a sequence of negotiation base lines. Agents can negotiate upon a base line with rather simple strategies. The experimental results show that near Pareto-efficient agreements can be reached effectively.
Proceedings of the fifteenth annual ACM symposium on Parallel algorithms and architectures - SPAA '03, 2003
We consider the practical problem of task assignment in a server farm, where each arriving job is... more We consider the practical problem of task assignment in a server farm, where each arriving job is immediately dispatched to a server in the farm. We look at the benefit of cycle stealing at the point of the dispatcher, where jobs normally destined for one machine may be routed to a different machine if it is idle. The analysis uses a technique which we refer to as dimensionality reduction via busy period transitions. Our analysis is approximate, but can be made as close to exact as desired, and is validated via simulation. Results show that the beneficiaries of the idle cycles can benefit unboundedly, due to an increase in their stability region, while the donors are only slightly penalized. These results still hold even when there is only one donor server and 20 beneficiary servers stealing its idle cycles.
23rd International Conference on Distributed Computing Systems, 2003. Proceedings., 2003
... prove to be very difficult: Many of the long-standing open questions in queueing theory invol... more ... prove to be very difficult: Many of the long-standing open questions in queueing theory involve the performance analysis of task assignment policies. ... consistent with validated stochas-tic models used to study a high-volume Web sites [10, 20], studies of scalable systems for ...
Proceedings of the fifteenth annual ACM symposium on Parallel algorithms and architectures - SPAA '03, 2003
We consider the practical problem of task assignment in a server farm, where each arriving job is... more We consider the practical problem of task assignment in a server farm, where each arriving job is immediately dispatched to a server in the farm. We look at the benefit of cycle stealing at the point of the dispatcher, where jobs normally destined for one machine may be routed to a different machine if it is idle. The analysis uses a technique which we refer to as dimensionality reduction via busy period transitions. Our analysis is approximate, but can be made as close to exact as desired, and is validated via simulation. Results show that the beneficiaries of the idle cycles can benefit unboundedly, due to an increase in their stability region, while the donors are only slightly penalized. These results still hold even when there is only one donor server and 20 beneficiary servers stealing its idle cycles.
In an electronic marketplace coalition formation allows buyers to enjoy a price discount for each... more In an electronic marketplace coalition formation allows buyers to enjoy a price discount for each item while combinatorial auction enables buyers to place bids for a bundle of items that are complementary. Coalition formation and combinatorial auction both help to improve the e ciency of a market and have received much attention from economists and computer scientists. But neither in laboratories nor in practice has there been literature on the situations where both coalition formation and combinatorial auctions exist. In this paper we consider an e-market where each buyer places a bid on a combination of items with a reservation cost, and sellers o er price discounts for each item based on volumes. We call coalition formation under this condition a Combinatorial Coalition Formation CCF problem since coalition formation is motivated by price discounts on single items while multiple items are complementary for buyers. By arti cially dividing the reservation cost of each buyer appropriately among the items we can construct optimal coalitions with respect to each item. We then try to make these coalitions satisfy the complementarity of the items, and thus induce the optimal solution. Based on this idea we present polynomial-time algorithms to nd a semi-optimal solution of CCF and a payo division scheme that is in the core of the coalition when linear price functions are applied, and in the pseudo-core when general price functions are applied. Simulation results show that the algorithms obtain solutions in a satisfactory ratio to the optimal value.
Proceedings of the 38th Annual Hawaii International Conference on System Sciences, 2005
We present a model for bilateral negotiations that considers the uncertain and dynamic outside op... more We present a model for bilateral negotiations that considers the uncertain and dynamic outside options. Outside options affect the negotiation strategies via their impact on the reservation price. The model is composed of three modules, single-threaded negotiations, synchronized multi-threaded negotiations, and dynamic multi-threaded negotiations. These three models embody increased sophistication and complexity. The single-threaded negotiation model provides negotiation strategies without specifically considering outside options. The model of synchronized multi-threaded negotiations builds on the single-threaded negotiation model and considers the presence of concurrently existing outside options. The model of dynamic multi-threaded negotiations expands the synchronized multi-threaded model by considering the uncertain outside options that may come dynamically in the future. Experimental analysis is provided to characterize the impact of outside options on the reservation price and thus on the negotiation strategy. The results show that the utility of a negotiator improves significantly if she considers outside options, and the average utility is higher when she both considers the concurrent outside options and foresees future options.
Multiagent and Grid Systems, 2008
Most of the prior work on multi-attribute negotiations usually assumes that agents have linear ad... more Most of the prior work on multi-attribute negotiations usually assumes that agents have linear additive utility functions on the attributes (e.g. [9, 11, 13]), or focuses on issue-by-issuenegotiations [1, 3]. The work that addresses complex utility functions and Pareto optimality, either assumes complete information, cooperative agents, the presence of a non-biased mediator [7], or makes the assumption that attributes have binary values [6,16,24]. This chapter aims to propose a model that can help agents negotiate multiple attributes with complex utility functions and incomplete information, but still maintains Pareto optimality. Thus, this chapter contributes to the existing literature in the following aspects: first, focusing on decentralized negotiations, our model can be applied to the situations with self-interested agents; second, the model allows agents to negotiate multiple attributes simultaneously and considers Pareto optimality of the negotiation outcome; third, the model can be applied to incomplete information scenarios where each agent knows neither the utility function nor the negotiation strategy of her opponent; and finally, the model can be applied to general situations where agents have complex utility functions in continuous negotiation domains. The rest of the chapter is organized as follows. Section 2 reviews the related work. Section 3 presents the model. Experimental analysis is provided in Sect. 4, and in Sect. 5 we conclude.
Electronic Commerce Research and Applications, 2004
We study the mechanism design problem of coalition formation and cost sharing in a group-buying e... more We study the mechanism design problem of coalition formation and cost sharing in a group-buying electronic marketplace, where buyers can form coalitions to take advantage of volume based discounts. The desirable mechanism properties include stability (in the core), and incentive compatibility with good efficiency. We show the impossibility to simultaneously satisfy efficiency, budget balance and individual rationality at a Bayesian–Nash equilibrium, and propose a mechanism in the core of the game. We then present and evaluate a group of reasonable mechanisms. Empirical results show positive correlation between stability and incentive compatibility (which is in turn related to efficiency).