Data underlying the publication: "Historical Watermark Matching: User-Guided System for Feature-based Recognition"

<strong>Data underlying the Publication "Historical Watermark Matching: User-Guided System for Feature-based Recognition"</strong><br>Watermarks in historical paper provide important evidence for tracing the provenance and production of written artifacts. However, existing digital archives rely primarily on semantic metadata, limiting accessibility and hindering systematic comparison. The paper "Historical Watermark Matching: User-Guided System for Feature-based Recognition" presents a user-guided recognition system for the syntax-based matching of historical watermarks. The system integrates automated image harmonization with interactive user input, allowing experts to refine denoising, binarization, and reconstruction processes for diverse watermark sources. Similarity is computed through features extracted from pre-trained convolutional networks and Zernike moments, enabling efficient retrieval of visually similar watermarks. Quantitative evaluations and qualitative studies with watermark experts demonstrate the system’s effectiveness, usability, and the immense potential that such a system has for improving the accessibility of watermark data. The system implementation and code are available at https://github.com/msmathcomp/historical_watermark_matching. This data set allows for the reproduction of the results from the paper.<br><strong>Organization of the data</strong>'db.pkl' The database used in both the quantitative and qualitative experiments of the paper. It encodes the features of the watermark images of the "Wasserzeichen-Informationssystem" (Watermark Information System, WZIS, https://www.wasserzeichen-online.de/. These features are used for automatic and user-guided comparison of the watermarks. In particular, the database allows for running the evaluation scripts within the repository https://github.com/msmathcomp/historical_watermark_matching.'UEQ_Data_Analysis_Tool_Version12.xlsx' The filled-in evaluation tool for the User Experience Questionnaire (UEQ) [1]. We have entered the survey results within the 'Data' sheet of the file. All subsequent sheets are then computing the results automatically [2]. Section 4.3 of the paper has a corresponding discussion of these results.<br>[1] Bettina Laugwitz, Theo Held, and Martin Schrepp. 2008. Construction and Evaluation of a User Experience Questionnaire. In HCI andUsability for Education and Work, Andreas Holzinger (Ed.). Springer Berlin Heidelberg, Berlin, Heidelberg, 63–76.[2] Martin Schrepp, Andreas Hinderks, and Jörg Thomaschewski. 2014. Applying the User Experience Questionnaire (UEQ) in DifferentEvaluation Scenarios. In Design, User Experience, and Usability. Theories, Methods, and Tools for Designing the User Experience, AaronMarcus (Ed.). Springer International Publishing, Cham, 383–392.

<strong>论文《历史水印匹配：基于特征识别的用户引导系统》所依托的数据集</strong><br>历史纸张中的水印可为纸质文物的溯源与制作研究提供重要佐证。然而当前主流的数字档案仅依赖语义元数据，不仅限制了数据的可访问性，也阻碍了系统性比对工作的开展。论文《历史水印匹配：基于特征识别的用户引导系统》提出了一套面向历史水印的基于结构匹配的用户引导式识别系统。该系统将自动化图像调和与交互式用户输入相结合，允许领域专家针对不同水印源数据优化去噪、二值化与重建流程。相似度通过从预训练卷积网络（convolutional networks）与泽尼克矩（Zernike moments）中提取的特征计算得出，可高效检索视觉相似的水印。针对水印领域专家开展的定量评估与定性研究证实了该系统的有效性与易用性，同时展现了此类系统在提升水印数据可访问性方面的巨大潜力。该系统的实现代码与程序可通过 https://github.com/msmathcomp/historical_watermark_matching 获取。本数据集可复现该论文中的全部研究结果。<br><strong>数据集组织形式</strong><br>'db.pkl'：本论文定量与定性实验中均使用的数据库。该数据库编码了「水印信息系统（Wasserzeichen-Informationssystem, WZIS，https://www.wasserzeichen-online.de/）」中水印图像的特征。这些特征可用于水印的自动比对与用户引导式比对。具体而言，该数据库可用于运行本仓库 https://github.com/msmathcomp/historical_watermark_matching 中的评估脚本。<br>'UEQ_Data_Analysis_Tool_Version12.xlsx'：用户体验问卷（User Experience Questionnaire, UEQ）[1]的填写完成版评估工具。我们已将调研结果录入该文件的「Data」工作表中，其余后续工作表将自动完成统计结果计算[2]。本论文第4.3节对该结果进行了对应讨论。<br>[1] Bettina Laugwitz、Theo Held 与 Martin Schrepp. 2008. 构建与评估用户体验问卷. 收录于《面向教育与工作的人机交互与可用性》, Andreas Holzinger 主编. 施普林格柏林海德堡, 柏林、海德堡, 第63–76页.<br>[2] Martin Schrepp、Andreas Hinderks 与 Jörg Thomaschewski. 2014. 在不同评估场景中应用用户体验问卷（UEQ）. 收录于《设计、用户体验与可用性：用户体验设计的理论、方法与工具》, Aaron Marcus 主编. 施普林格国际出版, 尚贝里, 第383–392页.