题目：Auctions and Negotiations: Models, Systems, and Behavioral Studies

时间：2016年10月12日（周三）下午2:00-4:30

地点：9号楼404室

主讲人：Gregory E. Kersten教授

主办单位：信工学院

主讲人简介

Dr. Kersten（Gregory E. Kersten, 康思腾），加拿大康考迪亚大学（Concordia University）约翰·摩森商学院 (John Molson School of Business)资深教授，意大利巴里理工大学机械工程与管理学院兼职教授，INFORMS群决策与谈判分会副主席及候任主席， SSCI Q1区期刊《Group Decision and Negotiation》主编。曾任意大利巴里理工大学、美国海军研究生院、香港科技大学、台湾国立中山大学、奥地利国际应用系统分析研究所等访问教授。长期从事单人与群决策、谈判分析、谈判支持、交换机制、拍卖、基于网络的系统开发、行为经济学等领域的研究。群决策与谈判系统的国际知名学者。研究项目多次获得加拿大自然科学与工程基金、加拿大社会与人文科学基金、德国洪堡基金、澳大利亚研究基金、加拿大贝尔基金、意大利CINECA基金等的支持。出版编辑9部专著/论文集，在权威SCI/SSCI学术期刊如Management Science, Decision Support Systems, European Journal of Operational Research等发表80余篇论文（Google Scholar引用4529次，h指数37）。

报告摘要

Auctions have become one of the key transaction mechanisms used in all kinds of transactions conducted on the web. Information and communication technologies (ICT) also contributed to the emergence of various types of auctions, including multi-attribute and combinatorial forward and reverse auctions.

Surveys of the government and business procurement managers show that over 90 per cent of procurement managers base their decisions on both price and non-price attributes (e.g., terms of payment and delivery, durability, lead-time, warranty, and service) [1]. In the past, multi-attribute procurement decisions had been made through a sequence of bilateral negotiations; procurement managers negotiated with one supplier at a time and made a decision to award the contract or engaged in negotiation with another supplier. In the past, multi-bilateral negotiations were possible but difficult and time- and effort-consuming, while simultaneous multi-bilateral negotiations were not possible, albeit with a few exceptions [2].

Auctions’ efficiency, savings and the leverage of the competition, on the one hand, and need for multi-attribute procurement transactions, on the other hand, led to the design and implementation of multi-attribute reverse auctions (MARA) [3]. One shortcoming of most MARAs is that they rely on the quasi-linearity assumption that can be met for the goods that were produced prior to an auction but not for those that are produced after the auction is completed.  In the latter cases, the relationship between the price and the costs causes that neither the transaction nor auction efficiency can be achieved [4]. Our theoretical results have been verified empirically with the use of our Imaras web-based system (http://invite.concordia.ca/imaras) in which the patented mechanism that does not require disclo-sure of the buyer’s preferences is implemented [5]; we found that when the efficient frontier is concave the bidders move away from the Nash-optimal solution.

Multi-bilateral negotiation (MBIN) is a mechanism that shares many similarities with MARA. ICT makes such negotiations not only a feasible but also a viable mechanism for organizations and consumers. We have designed and implemented MBIN system (http://invite.concordia.ca/imbins) and conducted multiple auctions and negotiation experiments. There are several interesting results of these studies, including the role and impact of concessions in auctions and negotiations [6], the impact of transparency on competition [7], and the efficiency of MARA and IMBIN mechanisms [8]. One result of these behavioral studies is a two-phase MARA followed by MBIN mechanism which has the purpose is to combine the efficiency of auctions in terms of time and effort with the solution efficiency of decision-supported negotiations [9].

A related stream of research involves bilateral negotiations and the analysis of the negotiations conducted by human agents on behalf of their principals. We have designed and implemented the Inspire system (http://invite.concordia.ca/inspire) and conducted negotiations with participants from over 80 countries. The collected data was used to study cultural differences, impact of the interface, language and counterpart behavior. Most recently, we have conducted sentiment analysis and the principal impartation of preferences.

The analysis of the results from both MBIN and the bilateral negotiations led us to design software agents and employ them in human-agent negotiations [10, 11].

参考文献

[1] Ferrin, B. G. and Plank, R. E. Total Cost of Ownership Models: An Exploratory Study. Journal of Supply Chain Management, 38, 3 (2002), 18-29.

[2] Subramanian, G. and Zeckhauser, R. ‘Negotiauctions’: Taking a Hybrid Approach to the Sale of High Value Assets. Negotiation, 8, 2 (2005), 4-6.

[3] Pham, L., Teich, J., Wallenius, H. and Wallenius, J. Multi-attribute Online Reverse Auctions: Recent Research Trends. European Journal of Operational Research, 242, 1 (2015), 1-9.

[4] Kersten, G. E. Multi-attribute Procurement Auctions: Efficiency and Social Welfare in Theory and Practice. Decision Analysis, 11, 4 (2014), 215–232.

[5] Kersten, G. E. and Wu, S. Multi-attribute auctioning method and system. PACIS, 2015.

[6] Kersten, G. E., Vahidov, R. and Gimon, D. Concession-making in multi-attribute auctions and multi-bilateral negotiations: Theory and experiments. Electronic Commerce Research and Applications, 12, 3 (2013), 166-180.

[7] Kersten, G. E., Wachowicz, T. and Kersten, M. Competition, Transparency, and Reciprocity: A Comparative Study of Auctions and Negotiations. Group Decision and Negotiation, 25, 4 (2016), 693-722.

[8] Yu, B., Kersten, G. E. and Vahidov, R. Traders’ Subjective Appraisals: Comparison of Negotiations and Auctions. Journal of Organizational Computing and Electronic Commerce, 25, 3 (2015), 233-261.

[9] Kersten, G. E. and Al-Basha, F. Beyond Dynamic Pricing: Dynamic Product Configuration with Auction/Negotiation Mechanisms. Luxemburg, 2015.

[10] Vahidov, R., Kersten, G. and Saade, R. An experimental study of software agent negotiations with humans. Decision Support Systems, 66 (2014), 135-145.

[11] Li, M., Vo, Q. B., Kowalczyk, R., Ossowski, S. and Kersten, G. Automated negotiation in open and distributed environments. Expert Systems with Applications, 40, 15 (2013), 6195-6212.