孙琳

〔摘 要〕[目的/意义]产品情报的有效获取能够帮助企业及时掌握市场动态及竞争态势，并有利于提升产品竞争力。对产品情报及其知识需求呈现跨领域、跨行业、多主体的综合性及深度挖掘商业价值的迫切性。[方法/过程]构建细粒度的产品情报知识元模型，通过基于证据理论的多属性融合过程，实现产品情报知识元属性集与关系集描述空间的扩展以及知识元属性关系网络的生成。[结果/结论]实例研究从业务和竞争力两个视角，实现产品情报知识元属性集的融合与深加工，为产品情报的隐性关联知识发现提供相对完备的知识基础。证明了基于证据理论的多属性融合方法对产品情报知识元融合的可行性和有效性。

〔关键词〕产品情报；多属性融合；知识元；证据理论；属性关系网络

DOI：10.3969/j.issn.1008-0821.2018.10.010

〔中图分类号〕N99 〔文献标识码〕A 〔文章编号〕1008-0821（2018）10-0065-08

〔Abstract〕[Purpose/Significance]Effective acquisition of product intelligence can help enterprises timely grasp market dynamics and competitive situation，as well as improve product competitiveness.Product intelligence and knowledge needs show not only the comprehensiveness of cross-domain，cross-industry and multi-subject，but also the urgency of deep exploration of business value.[Method/Process]A fine-grained knowledge element model of product intelligence was constructed.Through the multi-attribute fusion process based on evidence theory，the description space of the attribute-set and relation-set of product intelligence was extended，and the corresponding knowledge element attribute-relation network was generated.[Result/Conclusion]From two perspectives of business and competitiveness，the case study realized the fusion and deep processing of the attribute-set of product intelligence knowledge element，which provided a relatively complete knowledge base for the discovery of implicit association knowledge of product intelligence.It also proved the feasibility and effectiveness of multi-attribute fusion based on evidence theory for fusing product intelligence knowledge elements.

〔Key words〕product competitive intelligence；multi-attribute fusion；knowledge element；evidence theory；attribute-relation network

產品被视为企业成功的关键因素之一，产品情报是企业竞争情报分析与利用的重要领域。对产品情报的跟踪与融合还能够帮助决策者更加全面深入地了解市场需求动态，认清竞争形势，不断提升产品的竞争力。Amarouche K等[1]综述了现有产品竞争情报挖掘研究的主要研究成果；Mariadoss B J等[2]对销售人员的产品知识、竞争情报的行为和表现之间关系进行建模和测试；Xiao S等[3]提出了一种计量经济偏好度量模型用于从在线产品评论中提取总体消费者偏好。虽然大量的研究将产品竞争情报成功地应用于解决不同领域的商业决策，但几乎仅支持单一的应用领域，鲜有研究给出一个能够描述跨领域、跨行业、多主体的复杂情报要素的描述和综合解析，从而影响了产品情报辨识及其知识表示的通用性和全面性。……