知识共聚:领域分析视角下的知识聚合模式

Knowledge Copolymerization: A Knowledge Aggregation Mode Under the Perspective of Domain Analytic Paradigm

[目的／意义]当前知识聚合模式研究侧重“依据何种知识关联开展知识聚合”，本文补充性地探索“利用知识关联将知识单元聚合成何种形式”这一后续问题，以完善知识聚合模式研究和引导实践的深入。[方法／过程]借鉴化学领域中对聚合反应两大类型划分的方法，提出以“是否保留知识单元间及其关联间的差异性”为根据，将知识聚合划分为知识类聚和知识共聚，并探讨知识共聚的基本实现形式。[结果／结论]领域知识是知识共聚开展的基础；以文档、词语为基本知识单元粒度，以用户需求入口和聚合目标资源为维度，知识共聚可通过四种基本形式实现：基于分面导航、基于多维概念关联推荐、基于知识元链接、基于资源潜在关联发现。

[ Purpose/significance ] Current research on knowledge aggregation mode focuses on the types of knowledge connection that knowledge aggregation is based on. As an indispensable complementary, this paper explores the follow-up question of what form knowledge elements can be aggregated based on those connections,it aims at perfect the research of knowledge aggregation mode and guide related practice. [ Method/process ] Taking the classification of polymerization in chemistry, this paper put forward to divide knowledge aggregation mode into homopolymerization and copolymerization according to ＂whether or not remain the differences between knowledge units and their correlations＂, and then explored the basic realization forms of knowledge copolymerization. [ Result/conclusion] Domain knowledge is the basis of knowledge copolymerization. Utilizing documents and words as two basic knowledge elements, and considering users＇ demand portals and target knowledge elements as two dimensions, knowledge copolymerization can be achieved through the following four basic forms： based on faceted navigation, multidimensional recommendation based on conceptual relation, based on knowledge element linking, and knowledge discover based on the potential connection of resources.