Estimation of Geographic Relevance for Web Objects Using Probabilistic Models

Estimation of Geographic Relevance for Web Objects Using Probabilistic Models
复制标题

DOI:
10.1007/978-3-540-89903-7_12
复制
发表时间:
2008-12
期刊:
--
影响因子:
--
通讯作者:
Taro Tezuka;Hiroyuki Kondo;Katsumi Tanaka
Taro Tezuka;Hiroyuki Kondo;Katsumi Tanaka
中科院分区:
其他
文献类型:
--
作者:
Taro Tezuka;Hiroyuki Kondo;Katsumi Tanaka

文献摘要

相似文献

地理限制性网络搜索的迅速增加使得确定网络内容的地理相关性成为一项重要任务。我们已经开发了一种方法,使用高斯混合模型来估计网页和任意主题的地理相关性,并实现了一个可视化界面,将页面和主题映射到地理空间。该系统使用户能够检索与任意地理区域相关的网页和在Web上表达的主题。
The rapidly increasing use of geographically restrictive web search has made determination of the geographic relevance of web content an important task. We have developed a method that uses Gaussian mixture models to estimate the geographic relevance of web pages and arbitrary topics and have implemented a visualization interface that maps pages and topics to geographic space. The system enables the user to retrieve web pages and topics expressed on the Web that are relevant to an arbitrary geographic area.