用户需求导向的产品集成创新计算模型研究

讨论了现有创新模型在形式化、预测和反馈等方面存在的不足，提出一个根据需求进行筛选、从用户角度进行分析并引入反馈的新模型，用加权属性描述个体，根据大项集进行筛选，利用模糊聚类生成方案，根据评价结果校正属性关系。新模型基于属性关系蕴含知识的假设，具有较好的创新效果和自适应能力，主要用于集成创新。

Research on requirement-oriented product integration innovation model

Deficiencies of existing models were discussed at aspects of formalization, predication, feedback, etc. A new model was introduced, filtering by requirement information, analyzing from uses' point of view, and taking efficient feedback. Any product is formalized into attributes with weight, big item sets utilized to filter, fuzzy clustering leveraged to generate proposals, and estimation fed back to revise relation of attributes. The model supposes knowledge is implied in relations, takes on preferable innovation effect and adaptability, and primarily applies to integration innovation.