基于POI地标与路径相似度的IP定位研究  

IP GEOLOCATION BASED ON POI LANDMARKS AND PATH SIMILARITY

作　　者：金弋然 陈永乐 姚青桦 陈俊杰 Jin Yiran;Chen Yongle;Yao Qinghua;Chen Junjie(School of Information and Computer,Taiyuan University of Technology,Taiyuan 030024,Shanxi,China)

机构地区：[1]太原理工大学信息与计算机学院,山西太原030024

出　　处：《计算机应用与软件》2022年第10期77-86,共10页Computer Applications and Software

基　　金：山西省重点研发计划项目(201903D121121);山西省自然科学基金项目(201701D111002)。

摘　　要：由于缺少PlanetLab等公共地标平台,高精度IP定位技术仍然面临着巨大的挑战。为解决这一问题,提出一种基于大规模兴趣点(Point of Interest,POI)地标与路径相似度的IP定位算法,爬取和去噪数字地图上的POI数据并跟踪地标探测路径来收集地标信息。进一步设计了加权二维字符串子序列内核算法挖掘路由路径相似性特征,以及基于半径领域的离群点检测聚类算法来定位目标节点。该算法挖掘了北京市和上海市的2038个和1429个可用地标。实验表明,该算法显著提高了IP定位技术的稳定性与准确性,实现了5.7 km的平均定位误差。Due to lack of public landmark platforms like PlanetLab,high-precision IP geolocation still faces a great challenge.To solve this problem,this paper proposes an IP geolocation algorithm based on large-scale POI(point of interest)landmarks and path similarity.It collected landmark information by crawling and denoising POI data from digital map and tracking landmark detection path.Furthermore,this paper designed a weighted two-dimension string subsequence kernel algorithm to mine routing path similarity feature and circle-based outlier detection clustering algorithm to geolocate target node.This algorithm mined 2038 and 1429 available landmarks of Beijing and Shanghai respectively.The experiment shows that this algorithm obviously improves the stability and accuracy of IP geolocation,and achieves 5.7 km mean geolocation error.

关 键 词：IP定位 POI 地标 路径相似度 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]