基于创新点关联的自助查新模式研究

Self-help Novelty Retrieval Model Basing on Innovation Relevance

【目的/意义】目前的科技查新模式由于没有挖掘文献内容和关联发现机制,因此难以解决查新项目技术要点的定位及查新周期过长问题。【方法/过程】本文提出一种基于创新点关联发现的自助查新模式。该模式包括三个部分:挖掘科技文摘的创新点,将创新点分解为问题、方法、结果实体;将实体进一步分解为主谓宾语义关系;建立创新点语义知识库,实现在创新点语义知识库中自助查新。【结果/结论】实验结果表明,该模式易于定位查新点,具有了隐含关系发现,提高了查新的质量和查新速度。

【Purpose/significance】Due to the lack of literature content mining and associated discovery mechanisms, it isdifficult for the current sci-tech novelty retrieval model to solve the problem of locating the technical points of noveltysearch project and the problem of too long period of novelty retrieval.【Method/process】This paper proposes a self –helpnovelty retrieval model based on the innovation relevance,The model consists of three parts： mining the innovation pointsof the science abstracts, decomposing the innovation points into problems, methods, and result entities; further decomposingthe entities into the semantic relationships of the subject, predicate, and the objects; and establishing a semantic knowledgebase for innovative point to achieve self –help novelty retrieval in it.【Result/conclusion】The experimental results showthat this model is easy to locate new points, achieves the implicit relationship discovery, and improves the quality and speedof novelty retrieval.