摘要
现今,各行各业对科学技术奖励越发重视,所获奖励已成为衡量企业、大学科技实力的重要体现。在科技奖励的评审环节,人为打分的方式无法准确、客观的反映各个项目的真实优劣。其中,支撑申报项目的各项知识产权证明,是反映申报主体真实性、先进性的重要证据,但评审专家往往是特定技术领域的专家,受限于自身的专业知识结构、评审时间短而无法查阅所有知识产权等原因,难以对各项知识产权给出非常准确的判断。另外,目前评审过程仅利用计算机系统进行数据整理、统计、发布等,极少利用自动工具进行辅助打分。本文从知识产权本身出发,利用神经网络自动拟合能力,代替人工进行辅助评审,为人工评审提供参考。
Nowadays,all walks of life pay more and more attention to science and technology awards,and the rewards have become an important embodiment to measure the scientific and technological strength of enterprises and universities.In the judging process of science and technology awards,the way of scoring is not able to accurately and objectively reflect the true and inferior of each project.Among them,the various intellectual property certificates supporting the declared project are important evidences reflecting the authenticity and advancement of the reporting entity,but the reviewing experts are often experts in specific technical fields,limited by their own professional knowledge structure and short review time.It is difficult to give very accurate judgments on various intellectual property rights by consulting all intellectual property rights and other reasons.In addition,the current review process only uses computer systems for data collation,statistics,release,etc.,and rarely uses automated tools for auxiliary scoring.Based on the intellectual property itself,this paper uses the neural network automatic fitting ability instead of manual for auxiliary review,which provides reference for manual review.
作者
王晨辉
刘雯静
赵理
WANG Chen-hui;LIU Wen-jing;ZHAO Li(State Grid Information&Telecommunication Branch,BeiJing 100761,China)
出处
《新一代信息技术》
2019年第14期72-77,共6页
New Generation of Information Technology
基金
国家重点研发计划(项目编号:2018YFD0501103)。
关键词
科技奖励
知识产权
资产评估
BP神经网络
Technology reward
Intellectual property
Asset valuation
BP neural network