基于情感分析的altmetrics学术质量评价方法研究

［目的/意义］Altmetrics通过API跟踪和分析学术文献的网络活动，根据影响力分值评价学术成果的影响力。然而，其评价学术影响力不考察读者态度极性，仅提供简单的公众关注度信息，不能直接体现学术成果的质量。为此研究altmetrics评价学术质量的可行性方案。［方法/过程］将情感分析与altmetrics相结合，并以新闻媒体、社交网站、文献数据库、学术论坛的提及、评论、引用、讨论指标为例，探讨altmetrics量化指标与情感分析相结合的理论框架。［结果/结论］构建基于情感分析的altmetrics加权综合运算模型，该模型虽有情感分析技术实现难度较大、计算过程复杂且缺乏实证研究的不足，但作为评价学术质量的一种尝试，值得探讨。

Altmetrics Academic Quality Evaluation Method Based on Sentiment Analysis

Purpose/significance] Altmetrics tracks and analyzes the online activities of academic literatures with API, and evaluates the influence of outcomes according to influence scores. However, its evaluation of academic influence does not examine readers' attitudes and cannot directly reflect the quality of academic literatures, only shows public attentions. The paper aims to study the feasible plan of altmetrics in evaluating academic quality.Method/process] It combines the sentiment analysis with altmetrics, takes the indicators of mentions, comments, cites, discussions of news media, social networking sites, academic forums, Twitter and blogs as examples for examples, and discusses the theoretical framework of combing altmetrics quantitative indicators and emotion analysis.Result/conclusion] Altmetrics weighted synthetical operational model is constructed based on sentiment analysis and to evaluation academic quality. Its deficiencies include the difficult sentiment analysis technology, complex calculation process, and lack of empirical research. But as an attempt to evaluate the academic quality, it is worth to try.