基于媒体数据的知识产权热点主题识别  

Identification for the Hot Topics of Intellectual Property Based on Media Data

作　　者：林德明[1] 王宇开 杨中楷[1] LIN Deming;WANG Yukai;YANG Zhongkai(Institute of Science Studies and S&T Management,Dalian University of Technology,Dalian 116024)

机构地区：[1]大连理工大学科学学与科技管理研究所,大连116024

出　　处：《中国发明与专利》2022年第2期5-10,共6页China Invention & Patent

基　　金：国家社科基金重大项目“新形势下我国科技创新治理体系现代化研究”(20&ZD074)阶段性成果。

摘　　要：网络新闻的媒体数据中对特定主题的关注情况在一定程度上能够反映政策的影响力和执行力。选取中新网与知识产权相关的新闻文本、发布时间、新闻分类等数据作为研究对象,应用深度学习和三词多维共现的方法,深入研究近十年中国知识产权主题的热度、分布以及重点领域。结果表明:媒体数据反映了政策影响下的热点主题变化,知识产权的热点主题呈现波动的态势,其中对知识产权创造的关注高于运用和保护,这一特征在专利和商标中体现尤为明显。The attention to specific topics in media data of online news can reflect the influence and execution of policies to a certain extent. The texts, publication time, classification and other texts related to intellectual property in China News Network are selected as research objects. Furtherly, the deep learning and multi-dimensional co-occurrence method is applied to identify the hot topics, distribution of keywords,and key points. The results implicate that the media data of online news portray the transformation of hot topics under the effect of policies, and the trend of hot topics of intellectual property is fluctuated, the creation of intellectual property is more compelling than the utilization or protection, especially for the patens and brands.

关 键 词：知识产权 媒体数据 深度学习 主题分析 

分 类 号：G306[文化科学]