一种基于行人位置描述的道路模拟表达方法
发布时间：2018-10-08 21:59
【摘要】：人们在户外行走时,通常使用自然语言表达其对自身位置的感知,在计算机中如何将定性的自然语言位置描述转换为定量的图形语言是地理信息系统、人工智能等领域面临的挑战。该文提出一种基于行人位置描述的道路模拟表达方法,将自然语言位置描述转换为道路模拟地图。首先设计一组空间认知实验,探讨道路上的行人在户外如何使用自然语言描述当前位置;然后基于行人的每条自然语言位置描述估算该位置的候选区域,在候选区域中随机产生该位置的一对坐标值;最后将行人在路上经过的多个位置点坐标数据组合,采用直线拟合和概率拟合两种方法对该道路进行模拟表达。实验结果表明,该文方法能达到较好的模拟表达效果,为自然语言到图形语言的转换研究奠定了基础。
[Abstract]:When people walk outdoors, they usually use natural language to express their perception of their own position. How to convert qualitative description of natural language position into quantitative graphic language in computer is geographic information system (GIS). Challenges in areas such as artificial intelligence. In this paper, a road simulation representation method based on pedestrian location description is proposed, which converts natural language location description into road simulation map. First, a set of spatial cognitive experiments was designed to explore how pedestrians on the road use natural language to describe the current location outdoors, and then to describe the candidate area for estimating the location based on each natural language location of the pedestrian. A pair of coordinate values of the position are randomly generated in the candidate region. Finally, the coordinate data of multiple points of pedestrian passing on the road are combined, and the road is simulated by two methods: straight line fitting and probabilistic fitting. The experimental results show that this method can achieve a good performance of simulation and expression, which lays a foundation for the study of the transformation from natural language to graphic language.
【作者单位】： 南京师范大学地理科学学院;南京师范大学虚拟地理环境教育部重点实验室;江苏省地理信息资源开发与利用协同创新中心;
【基金】：国家自然科学基金项目“基于自然语言交互的地理空间信息认知与模拟表达研究”(41571382);国家自然科学基金项目“向量组合学习框架下基于依存混合树的中文语义解析研究”(61472191) 江苏省高校自然科学研究重大项目“基于自然语言交互的地图模型近似转换机制与可视化研究”(15KJA420001)
【分类号】：P208
本文编号：2258410
[Abstract]:When people walk outdoors, they usually use natural language to express their perception of their own position. How to convert qualitative description of natural language position into quantitative graphic language in computer is geographic information system (GIS). Challenges in areas such as artificial intelligence. In this paper, a road simulation representation method based on pedestrian location description is proposed, which converts natural language location description into road simulation map. First, a set of spatial cognitive experiments was designed to explore how pedestrians on the road use natural language to describe the current location outdoors, and then to describe the candidate area for estimating the location based on each natural language location of the pedestrian. A pair of coordinate values of the position are randomly generated in the candidate region. Finally, the coordinate data of multiple points of pedestrian passing on the road are combined, and the road is simulated by two methods: straight line fitting and probabilistic fitting. The experimental results show that this method can achieve a good performance of simulation and expression, which lays a foundation for the study of the transformation from natural language to graphic language.
【作者单位】： 南京师范大学地理科学学院;南京师范大学虚拟地理环境教育部重点实验室;江苏省地理信息资源开发与利用协同创新中心;
【基金】：国家自然科学基金项目“基于自然语言交互的地理空间信息认知与模拟表达研究”(41571382);国家自然科学基金项目“向量组合学习框架下基于依存混合树的中文语义解析研究”(61472191) 江苏省高校自然科学研究重大项目“基于自然语言交互的地图模型近似转换机制与可视化研究”(15KJA420001)
【分类号】：P208
【相似文献】
相关期刊论文 前4条
1 龙毅;张翎;胡雷地;闾国年;;移动GIS中语音与自然语言的应用模式探讨[J];测绘科学技术学报;2008年01期
2 张连蓬,储美华,刘国林,江涛;车载智能地理信息查询系统及其自然语言接口[J];现代测绘;2005年01期
3 张连蓬,刘国林,江涛,李云岭,季民;受限自然语言查询在GIS中的应用[J];测绘学院学报;2002年04期
4 明小娜;龙毅;钱程扬;张翎;;基于受限自然语言的GIS命令解析方法[J];地球信息科学学报;2009年02期
相关博士学位论文 前1条
1 朱少楠;面向地理场景的“文—图—景”转换方法研究[D];南京师范大学;2013年
，本文编号：2258410
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2258410.html
