【摘要】 提出一种面向专业领域的谱系型知识检索方法,分为数据预处理和知识检索两个阶段。在数据预处理阶段,首先抽取原始数据中的知识点及时间属性,然后按照时间切片方法对抽取出的知识点进行年代划分,最后统计知识点的出现次数和知识点间的共现次数;在知识检索阶段,根据与待检索知识点有共现关系的知识点的所在年代确定其与待检索知识点间的谱系关系,然后在时间轴上按年代进行排列,从而可视化展示知识点之间的发展脉络关系。这种新型知识图谱有助于协助专业领域的初学者从海量文献信息中抽取出有价值的知识点,并快速地掌握其所含的研究主题和发展脉络关系,克服了使用者由于缺乏专业背景知识而造成的检索效果不佳的问题。更多还原

【Abstract】 A genealogical knowledge retrieval method was proposed to the area of expertise.The retrieval methodology was divided into two stages.On data pre-processing stage,the knowledge concept and its time attributes from the raw data was extracted.And then the knowledge concepts were sliced into several era slots.Finally,the frequency of each knowledge concept occurrence and co-occurrence between knowledge concepts were counted.On knowledge retrieval stage,first all of the co-occurrence knowledge concepts with the retrieved keyword were arrived,and then their genealogical relations by era slots in the timeline panel were displayed.This visual analytics method will show the development of expertise skeleton to the viewers and assist the beginners of this area in extracting instrumental knowledge from massive information.It also helps users grasping the area of expertise accurately and effectively,especially to those rookies without the background knowledge.更多还原