Information Economy

The objective of this research project is to develop a mechanism design theory and solve various incentive problems including auctions. In the Internet, resources are distributed and participation is voluntary. These properties enable us to obtain huge amount of information from the Internet. However, it brings about problems such as fraud in Internet auctions. An auctioneer cannot directly control bidders’ behavior, i.e., an auctioneer cannot force bidders behave honestly/sincerely. A solution to overcome this situation is to devise the auction mechanism. If committing a fraud results in suffering a loss, i.e., honesty is the best policy for all bidders, any bidders are not willing to commit a fraud. Here, a challenge is how to design a mechanism that gives an appropriate incentive to participants and can satisfy the designer’s objective.

Mechanism design is an interdisciplinary area including computer science and economics (game theory). Game theory should be incorporated into the study of computer science. For example, we face a free-rider problem when designing a file sharing system in P2P network. On the other hand, computer science can contribute to actually employ the mechanisms designed by using game theory. For example, the FCC bandwidth auction mechanism is feasible if we consider the use of computer.

This field is very active. You can find many presentations related to an economic approach in AI and agent technology conferences. Why do not you join this project if you are interested in both of computer science and economics and want to contribute to design the future society by designing new mechanisms?

Auctions have been widely studied in the field of multiagent research to efficiently allocate resources. A major focus in auction studies is asymmetric information among sellers and buyers, i.e., how to elicit private information from participants and find an efficient allocation. Internet and agent-mediated electronic commerce involve uncertainty, which causes various types of problems. We have been studied uncertainty of a participant’s identity, uncertainty of auctioned goods, and auctions in dynamic environments.

Consider a tennis court assignment problem. There are multiple players contending for particular timeslots, with an auction determining the winner. A bidder’s valuation in fine weather may be different from that in rainy weather. However, obtaining the accurate weather forecast is difficult for all the players. A solution for dealing with this problem is to execute auctions whenever an event occurs and then to re-allocate resources. Re-allocating resources, however, may cause disutility. Moreover, it does not always provide an equilibrium strategy because it can be viewed as a sequential auction, which means that we cannot accurately predict what outcome will be obtained.

To solve this problem, we have proposed an auction protocol that allows bidders to declare the cost due to re-allocation and then decides an allocation based on this cost of re-allocation as well as the surplus obtained from the allocated resources themselves in the realized situation. We proved that a bidder’s truth telling is in equilibrium (honesty is the best policy) and that a socially efficient allocation is obtained in the proposed protocol.

Human and contents/services: we have started to study matching problems including humans and contents/services. Recently, consumer generated media such as Wikipedia have been attracting attention, although many of these contents are not sufficiently reliable if we consider an application in medical care or education. To create more reliable contents, a problem of intellectual property should be dealt with. For example, a content developed in the university may be restricted for non-profit use. That is, different content providers impose different constraints on its use of their contents. Finding appropriate contents and agreement become difficult if multiple contents is combined.

Human and human: forming a virtual team might be required to carry out a task. In such a human-human matching, accurately evaluating the value of combination, i.e., deciding the members of a team that maximize its productivity is difficult because factors of learning, growing and development exist. This research leads to design the future labor market.

Our approach: To solve the above problems, we try to extend the mechanism design theory for the market to that for the market and the community. We change the viewpoint from “control” to “co-creation” and establish the basic theory of dealing with intellectual property and incentive in the market and community. Based on the developed theory, we design various matching systems that facilitate the cycle of production, use, reproduction of contents/knowledge.

[Conferences]

Zhixing Huang, Yuhui Qiu, Shigeo Matsubara. Designing a Refundable Auction for Limited Capacity Suppliers. Third International Conference on Semantics Knowledge and Grid (SKG-2007), pp.104-109, 2007.

Ahlem Ben Hassine, Shigeo Matsubara, and Toru Ishida. A Constraint-based Approach to Horizontal Web Service Composition. The Fifth International Semantic Web Conference (ISWC-2006), pp.130-143, 2006.

Shigeo Matsubara. Auction in Dynamic Environments: Incorporating the Cost Caused by Re-allocation. The 4th International Joint Conference on Autonomous Agents and Multiagent Systems (AAMAS-2005), pp. 643-649, 2005.

Shigeo Matsubara. Trade of a Problem-solving Task. Proceedings of the 2nd International Conference on Autonomous Agents and Multi-Agent Systems (AAMAS-2003), pp. 257-264, 2003.

Shigeo Matsubara. Accelerating Information Revelation in Ascending-Bid Auctions - Avoiding Last Minute Bidding -. In Proceedings of the Third ACM Conference on Electronic Commerce (EC-01), pp. 29-37, 2001.

[Symposiums and Workshops]

[Journals]

Shigeo Matsubara. Auction in Dynamic Environments: Incorporating the Cost caused by Re-allocation. Journal of Information Processing in Japan, Vol. 47, No. 4, 2006. (In Japanese)

Shigeo Matsubara. Trading a Problem-solving Task. Transactions of the Japanese Society for Artificial Intelligence, Vol.18, No.5, pp. 269-277, 2003. (In Japanese)

Shigeo Matsubara and Makoto Yokoo. Defection-free exchange mechanisms based on an entry fee imposition. Artificial Intelligence Journal, Vol. 142, No. 2, pp. 265-286, 2002.

Shigeo Matsubara. Promoting Information Revelation in an Ascending-Bid Auction. Transactions of the Japanese Society for Artificial Intelligence, Vol.16, No.6, pp. 473-482, 2001. (In Japanese)

Shigeo Matsubara. A Collusion-free Adaptive Pricing Mechanism. Transactions of the Japanese Society for Artificial Intelligence, Vol.16, No.4, pp. 375-383, 2001. (In Japanese)

Shigeo Matsubara and Makoto Yokoo. Fraud-free Exchange Machanisms in Electronic Commerce. Journal of Japanese Society for Artificial Intelligence, Vol.15, No.5, pp. 912-921, 2000. (In Japanese)