基于中心向量的KNN分类方法在城市属性信息获取中的应用

Application of KNN Classification Method Based on Center Vector in Obtaining Urban Attribute Information

当前,利用非结构化数据进行信息提取是地理信息领域的研究热点。针对这一热点,本文提出了基于中心向量的KNN分类方法。首先,对向量空间模型的特征项选择及权重计算方法进行了改进;其次,利用KNN算法对非结构化数据进行了分类与排序;最后,以与郑州市相关的200篇网络资源为非结构数据语料库进行验证。实验表明:该方法能够有效地提取非结构数据的城市属性信息,计算效率较高,实用性较强。

At present,using unstructured data for information extraction is a hot topic in the field of geographic information.Aiming at this hot spot,this paper proposes a new classification method of KNN based on center vector.Firstly,the feature selection and weight calculation method of vector space are improved;On this basis,the unstructured data are classified and sorted by KNN algorithm;Finally,200 network resources related to Zhengzhou are verified as unstructured data corpus.The experimental results show that the method can extract the attribute information of unstructured data effectively and is efficient and practical.