基于大规模政府公文智能处理的知识发现及应用研究

Research on Construction and Application of a Knowledge Discovery System Based on Intelligent Processing of Large-scale Governmental Documents

政府公文是一种重要的知识资源。对公文资源的智能处理,实现公文中的知识发现,有利于促进公文的智能知识管理,提高政府公文在拟制、审批、传阅和归档等流转处理中的智能化水平和行政效能,对推进数字政府建设和提升政府治理能力都具有重要意义。但当前,鲜有面向我国政府公文智能处理和知识发现的相关研究。鉴于此,本文对该内容进行探索与研究。为实现大规模政府公文的智能处理,本文提出了政府公文的内容结构解析、主题自动标引、产生式自动摘要、重点内容提取与排序计算、面向政策/政令/行政执法文书的知识抽取与链接、基于公文的政令任务分解与责任对象自动匹配等处理流程及研究相关技术方法,并在此基础上构建知识发现系统和分析该系统的应用,也对其应用于特定类型公文中的知识发现进行了实例分析。

Governmental documents are an important form of knowledge resource. Knowledge discovery in governmental document resources based on intelligent processing can help achieve the goal of intelligent knowledge management and promote administrative efficiency in document fiction, document approval, document circulation, and document archiving. It can also significantly promote building digital government and enhance the effectiveness of government governance. To implement intelligent processing of large-scale governmental documents, this paper proposes a series of processing steps, including content structure analysis, subject automatic indexing, abstractive summarization, key content extraction and sorting, knowledge extraction and linking of policy/decree/administrative enforcement documents, task dissociation, and right-responsibility matching in governmental decree documents. Based on the above processing steps, this paper constructs the knowledge discovery system and analyzes its application. Finally, it also presents a case of knowledge discovery in specific governmental documents.