论文部分内容阅读
[目的/意义]从论文层面计量角度出发,针对不同科研人员阅读文献的需求,综合传统计量指标和新型社媒网络指标,提出核心文献价值测度与识别的思路与框架,进而为领域科学研究推进、科研管理服务开展提供一定的参考。[方法/过程]将核心文献划分为经典、热门和前沿三类,综合多重指标运用熵权法和主成分分析方法构建文献价值测度体系,并以人工智能领域为例进行实证研究。[结果/结论]实现了针对不同科研需求的人工智能领域核心文献识别与推荐,揭示了不同类型核心文献所属方向、所载期刊、所选指标、时序影响等方面的差异。[局限]纯粹客观赋权法存在权重设定偶然性和不确定性的局限,需进一步改进。
[Purpose / Significance] From the aspect of etiquette measurement, according to the needs of different researchers to read the literature, combining the traditional measurement indicators and the new social media network indicators, we put forward the thinking and framework of core document value measurement and identification, and then promote the field scientific research , Research management services to provide some reference. [Methods / Procedures] The core documents are divided into three categories: classic, hot and frontier. The index system is constructed based on multiple indicators using entropy method and principal component analysis. The field of artificial intelligence is used as an example for empirical research. [Results / Conclusion] The core document identification and recommendation in artificial intelligence for different scientific research needs were realized, and the differences among different types of core documents, the periodicals, selected indexes and timing influence were revealed. [Limitations] There is a limit to the mere objective weighting laws that set contingencies and uncertainties, which require further improvement.