【摘要】：Co-location模式挖掘是在空間對(duì)象集中找到位置上頻繁出現(xiàn)在一起的模式。例如,假設(shè)兩種范疇的商業(yè)行為在空間上通常會(huì)有鄰近關(guān)系,則它們很有可能會(huì)被定義為co-location模式；一些不同的生物會(huì)生長(zhǎng)在具有鄰近關(guān)系的區(qū)域內(nèi),它們很有可能具有co-location模式。co-location模式的發(fā)現(xiàn)在現(xiàn)實(shí)生活中具有非常重要的意義,比如廣告商會(huì)在人群聚集的地方放置特定廣告等等。 本文主要研究如何在GIS平臺(tái)上實(shí)現(xiàn)空間數(shù)據(jù)挖掘的co-location算法。首先簡(jiǎn)要介紹了關(guān)于數(shù)據(jù)挖掘、空間數(shù)據(jù)挖掘和空間關(guān)聯(lián)規(guī)則的一些基本概念,以及空間數(shù)據(jù)與傳統(tǒng)數(shù)據(jù)的區(qū)別。其次,詳細(xì)介紹了關(guān)于GIS空間數(shù)據(jù)挖掘技術(shù),包括GIS的工作原理和GIS挖掘的一些方法,為后面的開發(fā)奠定基礎(chǔ)。然后介紹了本文核心算法的有關(guān)概念以及算法的基本思想和執(zhí)行的過程,并對(duì)算法的數(shù)據(jù)結(jié)構(gòu)進(jìn)行了設(shè)計(jì)。接下來(lái),使用ArcGIS Engine組件和.NET平臺(tái)的結(jié)合,開發(fā)了co-location算法的挖掘模塊,通過嵌入式GIS的基本功能,為算法的執(zhí)行提取空間數(shù)據(jù)庫(kù)中的數(shù)據(jù),實(shí)現(xiàn)了較好的操作性和可視化的效果。 最后,使用了昆明市區(qū)的地圖為例,從該地圖的圖層中選擇了三個(gè)空間特征,即教育機(jī)構(gòu)、金融機(jī)構(gòu)和醫(yī)療衛(wèi)生機(jī)構(gòu),將它們通過所在行業(yè)的類型各自劃分為幾類進(jìn)行了分析。使用本文開發(fā)的系統(tǒng),通過設(shè)置不同的參數(shù),挖掘不同條件下的co-location規(guī)則,發(fā)現(xiàn)這些規(guī)則反映的實(shí)際情況,說(shuō)明該算法具有較強(qiáng)的實(shí)際應(yīng)用意義。
[Abstract]:Co-location pattern mining is a pattern that frequently appears together in the spatial object set. For example, assuming that the two categories of business behavior usually have adjacent relationships in space, they are likely to be defined as co-location patterns. Some different organisms will grow in adjacent areas, and they are likely to have co-location patterns. The discovery of co-location patterns is of great significance in real life. For example, advertising chambers place specific advertisements in places where people gather, and so on. This paper mainly studies how to implement the co-location algorithm of spatial data mining on GIS platform. Firstly, some basic concepts of data mining, spatial data mining and spatial association rules are briefly introduced, as well as the differences between spatial data and traditional data. Secondly, the GIS spatial data mining technology, including the working principle of GIS and some methods of GIS mining, is introduced in detail, which lays the foundation for the later development. Then the related concepts of the core algorithm, the basic idea and the execution process of the algorithm are introduced, and the data structure of the algorithm is designed. Next, using the combination of ArcGIS Engine component and. Net platform, the mining module of co-location algorithm is developed. through the basic function of embedded GIS, the data in spatial database is extracted for the execution of the algorithm. The good operation and visualization effect are realized. Finally, using the map of Kunming urban area as an example, three spatial characteristics are selected from the layer of the map, namely, educational institutions, financial institutions and medical and health institutions, and they are divided into several categories through the types of their industries. Using the system developed in this paper, by setting different parameters and mining co-location rules under different conditions, the actual situation reflected by these rules is found, which shows that the algorithm has strong practical application significance.
【學(xué)位授予單位】：云南大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2011
【分類號(hào)】：TP311.13

【參考文獻(xiàn)】

相關(guān)期刊論文 前6條

1 巨s，

本文編號(hào)：2481087