Evaluating the communication (interfaces) between decision makers and product recommendation agents in E-business to improve industrial productivity

因特网上提供的产品种类繁多，数量众多，因此需要决策支持系统来帮助用户选择“正确的”产品。长期研究计划的重点是改善促进决策者和基于网络的产品推荐代理(RAS)之间沟通的接口，这些RAS通过引出决策者的采购需求并相应地提出建议来提供帮助。其目的是使RAS更加个人化和了解用户的需求，更加灵活地使用，更加透明和值得信赖，以增加其在电子采购中的利用。这将提高采购者、产业工人和专业人员的生产力，实现网上采购对他们自己和他们的组织的好处。 除了减少信息过载和提高采购质量外，拟议的研究方案还具有重要意义，原因如下。首先，推荐技术的采用仍处于早期阶段。调查显示，寻找有关新供应来源的信息是互联网最重要的用途，因此，RAS对于通过互联网购买产品的个人和公司将特别有价值。其次，从业者已经意识到，RAS的用户界面对于让买家轻松发现新产品内容至关重要。因此，我们的目标是改进用户和RAS之间界面的输入端(例如，RA如何最好地引起用户对产品属性的偏好)，并增强界面的输出端(例如，RA的建议如何最好地呈现给用户)。第三，用户的信任是电子商务和RA利用的关键成功因素。本项目提出了通过使RAS的推理对用户更加透明和可信来增强用户信任的新方法。从业者可以通过重新设计他们的RA接口来从这项工作中受益，从而获得竞争优势并使购买决策更加有效。

The large variety and amounts of products available on the Internet have given rise to the need for decision support systems that can assist users in choosing the "right" products. The long-term research program focuses on improving the interfaces that facilitate the communication between decision-makers and web-based product Recommendation Agents (RAs) that provide assistance by eliciting the purchasing needs of decision-makers and making recommendations accordingly. The objective is to make RAs more personalized and understanding of users' needs, more flexible to use, and more transparent and trustworthy in order to increase their utilization in e-procurement. This will enhance the productivity of buyers, industrial workers and professionals by realizing the benefits of web-based procurement for themselves and their organizations. In addition to reducing information overload and improving procurement quality, the proposed program of studies is important for the following reasons. First, adoption of recommendation technology is still in its early stages. Surveys show that finding information about new sources of supply is the most important use of the Internet, therefore RAs would be especially valuable for individuals and firms purchasing products over the Internet. Second, practitioners have realized that the user interfaces to RAs are crucial to make it easy for buyers to discover new product content. As such, we aim to improve the input side of the interface between users and the RAs (e.g., how best does the RA elicit users' preferences of product attributes), and to enhance the output side of the interface (e.g., how an RA's advice be best presented to user). Third, users' trust is a key success factor for e-commerce and RA utilization. This project proposes new ways to enhance users' trust by making RAs' reasoning more transparent and credible to the users. Practitioners can benefit from this work by reengineering their RA interfaces both to gain competitive advantage and to make purchasing decision-making more effective.