文章关注公众科学分类标注项目,运用系统综述方法,在Web of Science核心合集中筛选出45篇文献作为证据来源,并根据EBLIP证据等级模型为证据进行评级,进而提出公众科学分类标注项目设计建议,即合理设置任务粒度,平衡数据质量与参与难度;适当设置"不知道"选项,平衡数据质量与分类效率;加入游戏化元素,提升参与过程的趣味性;广泛招募公众,重视但不能依赖"超级参与者";设置物质和精神激励,吸引公众加入并持续参与;采用通俗语言而非专业术语,关注非正式沟通渠道;提供教程指导,发挥科普作用;设置分类标注次数阈值,平衡数据质量和项目进度;以算法汇总参与者共识,获得最终分类结果.
Systematic Review and Evidence-Based Practice Strategies of the Design of Citizen Science Classification Projects in Library
This paper focuses on CS classification projects,employs a systematic review approach,selects 45 articles as sources of evidence from the Web of Science core collection,and it proposes recommendations design of citizen science classification projects:reasonably set the task granularity,balancing data quality and participation difficulty;appropriately set the"I dont know"option,balancing data quality and classification efficiency;implement gamification elements to make the process fun;recruit the public broadly,valuing but not relying on the"super-participants";set up material and spiritual incentives to attract the public to join and continue to contribute;use plain language rather than jargon and pay attention to informal communication channels;provide tutorial guidance and play a part in popularizing science;set the number of classifications each object gets,balancing data quality and project schedule;choose the algorithm to derive consensus from various labels,obtaining the final classification results.
Evidence-Based library information practiceEBLIPCitizen scienceEvidence ratingGLAMOpen research activitiesClassification projects
万方数据
维普
