Performance of online reputation mechanisms under the influence of different types of biases

Rating mechanisms are becoming an essential tool for online shoppers. There have been many reputation mechanisms proposed in the literature that try to mitigate the sometimes adverse effects of strategic behavior of the traders to alter their true reputation scores. At the same time, several studies have shown the presence of bias (often unintentional) in product ratings and their harmful effect on buyers and product sales. To date there has not been any work done that studies the performance of reputation mechanisms in the presence of various types of biases. This study is a first attempt to fill that gap. We present three reputation mechanism models that have been discussed in the literature, discuss three types of biases that have been studied in the past, and study the performance of the three reputation mechanisms in the presence of these biases by using a multi-agent system that simulates the interactions among buyers and sellers in an electronic marketplace. Using replicator dynamics we also show the effect of these models on the behavior of buyers and sellers as the buyers choose the type of reputation models they prefer to use and the sellers change their strategic behavior in response.     

E-commerce trust metrics and models

Traditional models of trust between vendors and buyers fall short of requirements for an electronic marketplace, where anonymous transactions cross territorial and legal boundaries as well as traditional value-chain structures. Alternative quantifications of trust may offer better evaluations of transaction risk in this environment. This article introduces a notion of quantifiable trust and then develops models that can use these metrics to verify e-commerce transactions in ways that might be able to satisfy the requirements of mutual trust. The article uses two examples in illustrating these concepts: one for an e-commerce printing enterprise and the other for Internet stock trading     

A regulated anonymous cryptocurrency with batch linkability

Cryptocurrencies such as Bitcoin use blockchain to conduct peer-to-peer value transmission. Nevertheless, the publicly nature of on-chain data might violate the privacy of the users. Subsequently, several anonymous cryptocurrencies, such as Zerocash and Monero, were proposed to enhance the privacy of cryptocurrencies. However, the strong privacy makes these cryptocurrencies perfect tools for illegal gains such as money laundering, extortion, and terrorist financing. As a result, regulation becomes a necessity for cryptocurrencies. In order to balance the contradiction between privacy and regulation in cryptocurrencies, in this paper, we propose a new regulated anonymous cryptocurrency protocol that can protect the privacy of honest payers while enabling a tracing authority to find out all the correlations among a batch of dubious transactions by a single query, and even trace malicious payers’ real identity if necessary. We formalize its system model and security model, including anonymity, sort-blindness, non-frameability and linkability. We also demonstrate that the proposed protocol achieves these desirable security properties with detailed security analysis. Finally, we show the validity and feasibility of this protocol by implementing a prototype system.     

考虑线上线下零售商竞争的制造商双渠道定价策略

根据现有的不同电商销售平台模式,将制造商双渠道销售模式划分为自营模式、平台模式和混合模式.分析在制造商主导的双渠道供应链中,3种不同渠道模式下价格竞争系数和佣金比例系数对其他经济变量的影响,以及制造商的最优定价决策与渠道选择问题.通过模型和数值分析发现,在平台模式和混合模式中,当制造商保证佣金比例系数在一定范围内,制造商利润随佣金比例系数的增加而减少,传统零售商、网络零售商和供应链总利润随佣金比例系数的增加而增加.在3种渠道模式中,混合模式对传统零售商和制造商利润增加更有利,平台模式对网络零售商和供应链总利润增加更有利.     

An incentive mechanism designed for e-marketplaces with limited inventory

In electronic marketplaces, reputation systems and incentive mechanisms are prevalently employed to promote the honesty of sellers and buyers. In this article, we focus on the scenario in which the inventory is in short supply, i.e. an e-marketplace with limited inventory (EMLI). The challenges are in twofold: (a) for sellers who aim to maximize their profit, they may intentionally conduct dishonest transactions since the limited products are likely to be sold out regardless of their reputation; (b) for buyers who intend to gain the limited products, they may provide untruthful ratings to mislead other buyers. To address these issues, we propose an incentive mechanism to promote buyer and seller honesty for this type of e-marketplaces. Specifically, the mechanism models the honesty of buyers and sellers as scores and reputation, respectively. It then offers a higher price to the products of more honest sellers (with higher reputation) and allocates the products to more honest buyers (with higher scores). In this way, both sellers and buyers are well encouraged to be honest. Furthermore, we impose proper membership fee on new sellers to cope with the whitewashing attack. We finally theoretically analyze and empirically demonstrate the efficacy of the proposed mechanism and its nice properties.     

Analyzing strategies of mobile agents on malicious cloud platform with Agent-Based Computational Economic Approach

Cloud computing is a service model that enables resource-limited mobile devices to remotely execute tasks from a server. The mobile agent is a software program installed in the mobile device to negotiate a diversity of commerce transactions with other mobile agents in the cloud. However, the negotiation plans carried by mobile agents are easily be eavesdropped on by the malicious cloud platforms, since the codes of mobile agents are read and executed by the cloud platform. Thus, sellers can tailor the negotiation plans to cheat buyers for seizing buyers’ profits in negotiations after eavesdropping on buyers’ negotiation plans. In this paper, we consider the buyers can take actions to resist the sellers’ cheatings, which is that the buyers can tailor their plans with extremely low demand to decrease sellers’ profits before migrate to the hosts. In this paper, we consider actions that buyers can take to resist sellers’ cheatings, that is the buyers can tailor their plans with extremely low demands before migrate to the cloud platform. Above situations between buyers and sellers are modeled as a mathematical model called Eavesdropping and Resistance of Negotiation (ERN) Game. The strategies of the buyers and sellers playing the ERN Game are analyzed by the Agent-Based Computational Economic approach. The simulation results show the cooperative strategies will be emerged between buyers and sellers in the ERN Game.     

The effect of intention analysis-based fraud detection systems in repeated supply Chain quality inspection: A context of learning and contract

As a result of the information asymmetry on product quality, there is a risk of unethical suppliers defrauding buyers in a supply chain. Buyers often conduct quality inspection on shipments and frame supply contracts to punish quality fraud. Due to cost concerns, buyers need to estimate the suppliers’ fraud possibilities and choose appropriate testing methods and frequencies. As suppliers’ fraud intentions depend on their cost-benefit analysis, it is possible to analyze suppliers’ fraud intention with appropriate modeling of their profit-seeking behavior. In this research, we are interested in how fraud intention analysis may affect the quality inspection process. It should be noted that quality inspection can be a repeated process, with suppliers and buyers conducting multiple rounds of transactions (including transactions with frauds) and learning about each other during the process. Their supply contracts may also affect suppliers’ profit-seeking attitude. We conduct a laboratory experiment to examine the effect of fraud intention analysis systems on inspection decision making considering the learning and contract effects. We put the experiment in the context of a dairy supply chain as a critical and interesting example application. The experiment shows that if there are no strong punitive terms for fraud in the contract, fraud intention analysis can improve buyers’ decision-making efficiency after controlling the learning effect, in terms of decision time, inspection cost, and correctness of rejecting suppliers’ fraudulent shipments.     

Analyzing the impact of intermediaries in electronic markets: an empirical investigation of online consumer-to-consumer (C2C) auctions

It has been widely believed that the Internet and electronic markets will eliminate traditional intermediaries. However, a close examination of the market mechanism indicates that the impact of intermediaries is indispensable. Specifically, we consider intermediaries’ impacts in price discovery and trust building. Intermediaries provide a buffer for temporary misalignment between supply and demand by buying low and selling high, and also help build trust by engaging in transactions with risk-averse buyers and sellers who otherwise will not participate in the market. Using a dataset from eBay’s online auctions, we examine empirically these two impacts of human intermediaries. We find that the presence of brokers has a significant impact on market liquidity, resulting in more successful trades and higher auction prices. In addition, we find that brokers are more likely to engage in transactions with less established sellers. Their presence reduces reputation penalty faced by these players and further facilitates the successful sale of items in auctions.     

Intelligent agents in electronic markets for information goods: customization, preference revelation and pricing

Electronic commerce has enabled the use of intelligent agent technologies that can evaluate buyers, customize products, and price in real-time. Our model of an electronic market with customizable products analyzes the pricing, profitability and welfare implications of agent-based technologies that price dynamically based on product preference information revealed by consumers. We find that in making the trade-off between better prices and better customization, consumers invariably choose less-than-ideal products. Furthermore, this trade-off has a higher impact on buyers on the higher end of the market and causes a transfer of consumer surplus towards buyers with a lower willingness to pay. As buyers adjust their product choices in response to better demand agent technologies, seller revenues decrease since the gains from better buyer information are dominated by the lowering of the total value created from the transactions. We study the strategic and welfare implications of these findings, and discuss managerial and technology development guidelines.     

基于机器学习的比特币去匿名化方法研究

比特币是一种去中心化的匿名加密货币,是目前使用最广泛的数字资产之一,具有匿名性、无主权、无地域限制的特点,匿名性的特性也使得比特币被广泛应用于各种犯罪活动。为实现比特币的去匿名化,提出一种联合特征构造方法并构建随机森林与Softmax相结合的分类模型。为更好地区分不同类型比特币的交易行为,引用交易实体的概念,按照联合特征构造方法分别从地址、实体与交易网络结构3个方面在海量的交易数据中构造特征,并将其整合成联合特征向量。实验结果表明,该实体分类模型的类别识别精确率超过0.92,其能够有效提升执法机构对虚拟货币犯罪行为的调查取证能力。     

The concept of decentralized and secure electronic marketplace

For commerce (electronic or traditional) to be effective, there must be a degree of trust between buyers and sellers. In traditional commerce, this kind of trust is based on such things as societal laws and customs, and on the intuition people tend to develop about each other during interpersonal interactions. The trustworthiness of these factors is based, to a large extent, on the geographical proximity between buyers and sellers. But this proximity is lost in e-commerce. In conventional electronic marketplaces the trust among participants is supported by a central server which imposes certain trading rules on all transactions. But such centralized marketplaces have serious drawbacks, among them: lack of scalability, and high cost. In this paper we propose the concept of Decentralized Electronic Marketplace (DEM) which allow buyers and sellers to engage in commercial transactions, subject to an explicitly stated set of trading rules, called the law of this marketplace—which they can trust to be observed by their trading partners. This trust is due to a decentralized, and thus scalable, mechanism that enforces the stated law of the DEM. We implement an electronic marketplace for airline tickets in order to illustrate the feasibility of the proposed concepts for decentralized and secure electronic marketplace.     

Reconciling predictive and interpretable performance in repeat buyer prediction via model distillation and heterogeneous classifiers fusion

Repeat buyer prediction is crucial for e-commerce companies to enhance their customer services and product sales. In particular, being aware of which factors or rules drive repeat purchases is as significant as knowing the outcomes of predictions in the business field. Therefore, an interpretable model with excellent prediction performance is required. Many classifiers, such as the multilayer perceptron, have exceptional predictive abilities but lack model interpretability. Tree-based models possess interpretability; however, their predictive performances usually cannot achieve high levels. Based on these observations, we design an approach to balance the predictive and interpretable performance of a decision tree with model distillation and heterogeneous classifier fusion. Specifically, we first train multiple heterogeneous classifiers and integrate them through diverse combination operators. Then, classifier combination plays the role of teacher model. Subsequently, soft targets are obtained from the teacher and guide training of the decision tree. A real-world repeat buyer prediction dataset is utilized in this paper, and we adopt features with respect to three aspects: users, merchants, and user–merchant pairs. Our experimental results show that the accuracy and AUC of the decision tree are both improved, and we provide model interpretations of three aspects.

     

Modeling and optimization of two-stage procurement in dual-channel supply chain

An online marketing platform should be designed to fairly take the benefits of buyers and suppliers into consideration based on their risk preferences and business strategies. In this paper, the dual-channel supply chain models are developed to implement the risk-averse strategy for buyers and risk-neutral strategy for suppliers, respectively. The buyers under the consideration are the manufacturers who acquire raw materials, parts, or components to make their final products. The major factors in the developed models include the risk preferences of buyers and suppliers, random price fluctuations of goods, and varying demands of final products. To reflect the purchasing practice of a manufacturer, (1) a supply chain is considered to have two supply channels, i.e., contract-based purchase with a lead-time before the goods are used and a direct purchase from online spot markets when the goods are used; (2) the time factor on decision making is specially taken into account, and the procurements are divided into the contract stage of purchase and online stage of purchase. Gaming analysis is conducted to develop the supply chain models for manufactures and suppliers to implement their purchasing or pricing strategies. The simulation is conducted and the result has shown that two-stage purchases in a dual-channel supply chain have improved the performances of suppliers and manufacturers in terms of the profits they can make, their supply–demand relations, and their adjustability to uncertainties in globalized and segmented markets. The proposed model has its significance for manufacturers to better control the price risk of goods and the demand risk of final products; on the other hand, suppliers can benefit from adjusting dynamic sales using online spot markets.     

Managing diagrammatic models with different perspectives on product information

In the context of engineering-oriented companies, a basic distinction can be made between the information needed during sales, engineering and manufacturing. When representing such information in diagrammatic models, these models will most often include both individual and shared content. For instance, during the sales phase, information about sales prices may be relevant, while this information is not relevant in the engineering and manufacturing phases, where, on the other hand, more detailed information about components and assembly is needed. However, such information models often share basic component definitions. Having this overlapping information across models means that when maintaining these, redundant work has to be carried out for the overlapping parts of the information. This, obviously, can be both time-consuming and a significant source of errors. In this paper, a modelling method for the management of diagrammatic models with different perspectives on product information is proposed. The aim of this method is to avoid redundant information across models, in order to reduce time needed for ensuring of consistency across models and minimize the chances of errors. The need for information models with different perspectives on the same product information is common in cases that involve design/redesign of complex products and/or the construction of product configurators. Besides a knowledge representation technique, the paper suggests two different solutions for software support, of which the application of one of these in an actual project is described.     

Customer satisfaction incentives with budget constraints

We study the design of incentive contracts based on customer satisfaction (CS) surveys with reward budget limits. We extend principal‐agent models to consider budget constraints, survey response rates, and correlation between CS measure and demand. We derive the optimal incentive contract and study the impacts of these factors on contract performance. In contrast to the common belief that customer future values are the drivers of CS incentives, we show that CS incentives can benefit principals even in a single‐period setting where customers bring no future value. Improvements can be achieved without increasing total reward, because the CS incentive program reveals additional information about agents' service effort and diversifies their risk. Such effects are overlooked in existing CS research. With consideration of correlation between sales and CS measures, we provide a metrics selection rule regarding which reward(s)—CS, sales commission, or both—should be included in an incentive plan. We also study cumulative incentive schemes based on commonly used average CS measures and show that such incentive schemes may fail to motivate agents to increase service effort. Therefore, designing proper reward schemes is a critical issue for effective CS management and deserves future research.     

The design and evaluation of an intelligent sales agent for online persuasion and negotiation

Online purchases from e-stores are getting popular among Internet users. Although many e-commerce activities, such as auctions, bargaining, and recommendation services are available, most e-stores lack clerk-like mechanisms to persuade potential buyers into buying products and to bargain with them for making a good deal. The objectives of this research are to design a lab prototype of a sales agent with persuasion and negotiation capabilities and to evaluate its effectiveness as a virtual clerk in an e-store. The prototypical intelligent sales agent (ISA) is equipped with reinforcement learning capabilities and an abstract argument framework. We conduct both laboratory and Internet experiments to assess ISA’s performance. The experimental results reveal that an e-store embedded within such a sales agent can improve a seller’s surplus and increase a buyer’s product valuation, willingness-to-pay, and satisfaction with the e-store.     

Dynamic adaptation of online ensembles for drifting data streams

The success of data stream mining techniques has allowed decision makers to analyze their data in multiple domains, ranging from monitoring network intrusion to financial markets analysis and online sales transactions exploration. Specifically, online ensembles that construct accurate models against drifting data streams have been developed. Recently, there has been a surge in interest in mobile (or so-called pocket) data stream mining, aiming to construct near real-time models for data stream mining applications that run on mobile devices. In such a setting, it follows that the computational resources are limited and that there is a need to adapt analytics to map the resource usage requirements. Consequently, the resultant models should not only be highly accurate, but they should also adapt swiftly to changes. In addition, the data mining techniques should be fast, scalable, and efficient in terms of resource allocation. It then becomes important to consider Return on Investment (ROI) issues such as storage requirements and memory utilization. This paper introduces the Adaptive Ensemble Size (AES) algorithm, an extension of the Online Bagging method, to address these issues. Our AES method dynamically adapts the sizes of ensembles, based on ROI usage patterns. We illustrate our approach by analyzing the performances against both synthetic and real-world data streams. The results, when comparing our AES algorithm with the state-of-the-art, indicate that we are able to obtain a high Return on Investment (ROI) and to swiftly adapt to change, without compromising on the predictive accuracy.     

考虑供应商保鲜努力的生鲜电商销售模式选择

基于一个生鲜供应商和一个生鲜电商组成的供应链,考虑生鲜供应商的保鲜努力对产品新鲜度的影响,建立转售模式、代理模式和混合模式下的博弈模型,研究生鲜供应商提供保鲜努力下生鲜电商的最优销售模式选择,分析销售模式差异对保鲜努力水平和销售价格的影响.研究结果显示:当佣金率较高且消费者对产品新鲜度比较敏感时,生鲜电商的最优销售模式是混合模式;否则,生鲜电商的最优销售模式是转售模式,而生鲜供应商的利润和供应链总绩效始终在混合模式下最大.生鲜供应商在代理模式和混合模式下的保鲜努力水平始终相等,佣金率较低时,代理模式和混合模式下的保鲜努力水平高于转售模式下的保鲜努力水平.此外,代理模式和混合模式下的销售价格随佣金率的降低而提高,佣金率低于一定水平时,混合模式下转售渠道的销售价格最高.     

Value-cell bar charts for visualizing large transaction data sets

One of the common problems businesses need to solve is how to use large volumes of sales histories, Web transactions, and other data to understand the behavior of their customers and increase their revenues. Bar charts are widely used for daily analysis, but only show highly aggregated data. Users often need to visualize detailed multidimensional information reflecting the health of their businesses. In this paper, we propose an innovative visualization solution based on the use of value cells within bar charts to represent business metrics. The value of a transaction can be discretized into one or multiple cells: high-value transactions are mapped to multiple value cells, whereas many small-value transactions are combined into one cell. With value-cell bar charts, users can 1) visualize transaction value distributions and correlations, 2) identify high-value transactions and outliers at a glance, and 3) instantly display values at the transaction record level. Value-cell bar charts have been applied with success to different sales and IT service usage applications, demonstrating the benefits of the technique over traditional charting techniques. A comparison with two variants of the well-known Tree map technique and our earlier work on pixel bar charts is also included.