随着知识经济和大学科时代的到来,构建跨学科专家团队势在必行,但目前构建程序尚不客观,同时也缺乏程式化的团队需求描述体系及专家可信度评价机制,为此我们提出了基于模糊匹配的跨学科专家团队发现算法。算法通过引入程式化逻辑符号,率先完成了团队需求向团队特征向量的转化,生成了基于知识智能的专家特征向量。然后,在特征词关联矩阵构建的基础之上,实现了两类向量的模糊匹配。最后,以P@N为评价指标,参考数学中收敛原理,设计了候选专家可信度评价程序。实例分析表明,该算法相比之前向量余弦运算方式更为准确可靠。
With the arrival of the era of knowledge economy and integration among disciplines, it is imperative to build interdisciplinary team of experts, but now the building process is not objective, meaningwhile, it is also lack of stylized description about team needs and credibility evaluation mechanisms for expert. In view of those, we present an interdisciplinary team of experts discovery algorithm based on the fuzzy matching. Firstly, we introduce stylized logical symbols to complete the transformation from team needs to team eigenveetor and generate the expert eigenvector based on knowledge intelligence. Thirdly, we achieve the fuzzy matching process on the basis of construction of keywords connection matrix. Finally, we design the credibility assessment procedure for candidate expert, which sets P@ N as evaluation indicator and refers to the mathematical principle of convergence. Instance analysis has demonstrated that the algorithm is more accurate and reliable than the previous eigenvector matching via cosine operations.