论文部分内容阅读
【目的】通过对专题知识库中文本资源的可视化展现为用户提供更直观的导航。【方法】在多层次文本聚类生成的资源划分结果的基础上,通过主题发现、降维处理与可视化展现等步骤,实现专题知识库中文本资源的可视化导航。【结果】提出一种TF-ICF主题词抽取算法,并综合利用优化的树图与散点图实现专题知识库的可视化展现,帮助用户便捷地了解知识库概况、定位所需关注的主题、理清各资源间的关联。【局限】在可视化展现过程中存在部分人工干预,知识库可视化展现的交互性仍有待改善。【结论】提出的可视化方法能较好地应用于专题知识库的资源展现,对进一步优化专题知识库的用户体验有重要意义。
[Purpose] To provide users with more intuitive navigation through the visual display of text resources in the topic knowledge base. 【Method】 On the basis of the results of resource division generated by multi-level text clustering, the visual navigation of the text resources in the topic knowledge base is realized through the steps of theme discovery, dimension reduction and visualization. 【Result】 A TF-ICF keyword extraction algorithm was proposed. The optimized tree and scatter chart were combined to realize the visual representation of thematic knowledge base, which helps users to easily understand the knowledge base and locate the topics of concern. Clear the link between resources. [Limitations] There are some human interventions in the process of visualization, and the interactive display of knowledge base visualization still needs to be improved. 【Conclusion】 The proposed visualization method can be well applied to the resource presentation of the topic knowledge base, which is of great significance to further optimize the user experience of the topic knowledge base.