政府网站信息资源知识元模型与可视化表征研究

Study on Knowledge Element Model and Visual Representation of Government Website Information Resources

[目的/意义]利用知识元模型理论研究政府网站知识服务效果的优化路径,辅以可视化表征技术,以降低大数据环境下政务用户信息获取的操作负载和知识加工的认知负荷。[方法/过程]依据相关知识元模型研究推理出符合政府网站信息资源属性特征的六元组知识元表示方法和四元组知识元本体结构,采用TextRank与HDP算法分别抽取政府网站信息资源关键词和主题词,并由领域专家根据抽取结果确定知识元,构建包含知识元本体库生成和可视化知识服务的政府网站信息资源领域知识元可视化表征模型。[结果/结论]通过政府网站发布的共享单车实例检验知识元可视化表征模型的有效性和可行性,为实现政府网站粗粒度信息服务转向以知识元为单位的细粒度知识服务范式开辟了新的研究思路,可视化知识服务模式增强了政务信息导航的结构化和用户解读领域文本语义的效果。

[ Purpose/significance ] The theory of knowledge element model is used to study the optimization path of the knowledge service effect of government Website, and the visual representation technology is helpful to reduce the operating load and the cognitive load of the information processing of the government users under the big data environment. [Method/process] According to the related knowledge element model, the six-tuple knowledge element representation method and the knowledge element ontology four tuple structures are deduced, which conforms to the characteristics of the information resources of the government Website. The TextRank and HDP algorithms are used to extract the key words and the subject words of the government Website information resources, and the domain experts determine the knowledge ac- cording to the extraction results. A visual representation model of government Website information resources knowledge el- ement is constructed, which includes knowledge element ontology database generation and visual knowledge service. [ Resnit/condusion] The shared bicycle as an example issued by the government Website tests the effectiveness and feasibility of the visual representation model of knowledge element, and it provides a new research idea for the transition from government Website document＇ s coarse-grained service to the knowledge element as a unit of fine-grained service, also with the help of visual knowledge services, the structured navigation of government information and the effect of user interpretation of domain text semantics are enhanced.