基于认知实验的地图汉字注记配置规则及其形式化表达研究

This research focuses on some basic Chinese cartographic annotation rules and relevant quantitative map labeling metrics in cartographic/GIS softwares. Cartographic annotation rules are mainly based on the working experience of some outstanding cartographers and most of them are represented by qualitative ways at the present time. .One of the major advantages of these ways is making the rules to be interpreted as simple as possible, but a lot of cartographic annotation rules will be easily affected by various subjective and objective factors originated from cartographers or their different working purposes at the same time. More importantly, it is hard to transform the qualitative cartographic annotation rules into quantitative metrics in cartographic/GIS softwares precisely and appropriately. .The most probable explanation is that it is difficult to explain to cartographers the logic behind what is considered to be "normal" or "acceptable" annotation rules. Some researches have shown that psychology theories(e.g. Gestalt theories) have had an immense influence on cartographic annotation rules, but the research methods are primitive, and the results are not comprehensive. .For reasons of improving cartographic annotation rules and representing them precisely, this research raises a new idea that we can reveal the logic of cartographic annotation rules based on geo-spatial cognition theories both from professional cartographers and general map readers. .The objects of this research are some basic Chinese cartographic annotation rules, and the technical means include geo-spatial cognition experiments, as well as knowledge maps in geographic knowledge visualization. The results mainly involve the quantitative and qualitative representation methods of characteristics in Chinese cartographic annotation rules, as well as some important spatial relationships between annotaions and annotations-features in different layers or scales. .The purposes of this research are trying to harmonize the crucial distinction between geo-spatial cognitive reasoning and automated map labeling algorithms design effctively, construct the sets of Chinese cartographic annotaion rules which can fit geo-spatial cognition principles, suit digital cartographic environment, and reflect artistic senses. As one of typical cases applying geo-spatial cognitive reasoning to improve map design quality, with the help of this research, we can clarify the geo-spatial cognitive logical meanings behind some qualitative working experience of cartographers, and provide some academic foundations of achieving an intelligent, automatd geo-information system.

现有地图注记配置规则主要来自制图者的工作经验，由于对其机制缺乏系统研究，制图时这些规则不仅极易受各种主客观因素影响而且难以实现精确化和定量化。研究表明，从心理认知角度可以初步解释注记配置规则的基本内涵，但方法简单原始，结论缺乏代表性和系统性。为此，本项目以地图汉字注记配置规则为研究对象，依据空间认知揭示认知过程和结果的研究进展，提出了从"制图者认知"和"读图者感知"角度共同揭示地图注记配置规则机制的科学设想。应用认知实验和地学知识可视化技术，深入研究地图汉字注记配置规则的心象地图特征、空间关系模型以及形式化表达技术三方面内容，实现地图汉字注记配置规则认知推理和自动化注记配置算法的有效融合，建立符合认知规则、适宜定量表达、具备美观特性的地图汉字注记配置规则知识库。主要目标是较为系统、完整地阐明地图注记配置规则空间认知机制，以便为实现基于朴素地理学的智能化GIS设计提供理论依据和典型研究范例。

地图注记配置规则是实现注记自动化配置算法设计和质量评价的重要依据。由于目前国内外地图注记配置规则主要来自制图者的实际经验，通常借助自然语言描述表达，因此，人们对注记配置规则的理解与应用不仅易受制图者个人素养和表现风格差异的影响，而且在数字制图环境下难以准确对其进行定量表达与模拟。更重要的是，现有配置规则大多基于制图者的单方面推理决策，较少顾及读图者对注记配置规则的认知机制，使得现有规则难以同时体现读图、制图两类主体在空间认知水平上的明显差异，进而影响制图结果的有效表达。. 为此，本项目主要研究内容围绕以下三方面展开：.（1）基于认知实验I的地图汉字注记配置规则心象地图特征；.（2）基于认知实验II的地图汉字注记配置规则空间关系特征；.（3）地图汉字注记配置规则的表达方法。. 研究结果表明：. （1）有衬线的仿宋体，因其强调了注记笔画的开始和结束，易读性比较高，适合图面要素数量较多时使用。这一结论与现有图式标准基本保持一致（现有标准中规定，乡镇及以上政府驻地汉字注记均采用等线体注出，一般村庄均采用仿宋体注出）。 . （2）针对字体+字大组合，本研究结果与现有图式标准的规定并未能保持一致。研究结果以仿宋体（高等级地物，有衬线字体，较大字大）+ 等线体（低等级地物，无衬线字体，较小字大）组合为最优，等线体（高等级地物，较大字大）+ 仿宋体（低等级地物，较小字大）组合为最差。虽然我们初步分析后者原因可能为该种组合清晰性、易读性不及前者，但仍需在后续研究中进一步加以科学证实。 .（3）针对点状要素注记优先级的约定，本研究结果与现有图式标准的规定并未能保持一致。首先，对于具有读图/制图经验被试（教师-制图者被试）而言，点状居民地注记优先级所承载心象地图信息量差异并不显著，基本可以认为此类被试对注记的优先级变化没有特别偏好。因此，从专业制图者角度来看，在点状要素注记配置中是否有必要强调其优先级顺序，值得在后续研究中进一步加以科学论证。其次，对于不具有读图/制图经验被试（学生被试），尽管点状居民地注记优先级所承载心象地图信息量差异显著，但本项目研究结果却与目前图式标准及前人研究经验有明显差别（特别是右侧、上侧位置），由此可见开展本项目相关研究具有特别意义，该优先级差异是否具有代表性及其产生原因仍有待后续研究中进一步加以科学论证。