基于TF-IDF算法的公文用户画像全文替换

Official Document User Portrait Based on TF-IDF Algorithm

政府公文是一种重要的知识资源。本文基于用户画像,实现了用户待办公文的有效排序及推荐。参考历史公文信息,利用TF-IDF算法获取用户画像中的公文关键词,与历史公文的文本相关性,实现待办理公文的推荐。基于结构熵权法和模糊分析,对影响公文处理效率的主客观因素提取权重。最终将用户画像与待办公文的主客观因素结合评估分数,实现待办公文合理排序。

Governmental official document is an important knowledge resource. Based on user portrait,this paper realizes the effective ordering and recommendation of user backlog documents. Referring to the information of historical documents,TF-IDF algorithm is used to obtain the key words of official documents in user portraits,and the text correlation between them and the historical documents is realized to recommend the documents to be handled. Based on structural entropy weight method and fuzzy analysis,the weights of subjective and objective factors affecting the efficiency of document processing are extracted. Finally,the user portrait is combined with the subjective and objective factors of the backlog documents to achieve a reasonable order of the backlog documents.