23.txt

This complete matlab for neural network

发信人: GzLi (笑梨), 信区: DataMining
标  题: [合集]什么软件可以得到一个网站的所有链接关系
发信站: 南京大学小百合站 (Wed Sep 11 12:44:55 2002), 站内信件

carantion (康乃馨) 于Sun Sep  8 10:58:13 2002提到：

就是从主页开始一直根据页面上的链接关系

得到整个网站的所有链接关系，也就是这个

网站的topo结构

谁知道有什么软件可以做到这样的功能？

哪里可以下载到这样的软件？

谢谢先


helloboy (hello) 于Sun Sep  8 16:15:57 2002提到：

I have tried this before.
But I have lost confidence because it's hard to extract linkage from dynamic
web pages. In fact, perhaps no linkage exist in origin web page because link
can be created dynamically.


kdd (kdd) 于Sun Sep  8 17:11:38 2002)
提到：

你可以用离线浏览器下载某个网站的所有页面，它分级目录存储，不过确实不能下载动态
页面。你看着办吧。。。


helloboy (hello) 于Sun Sep  8 18:37:52 2002提到：

It's slow. And I don't how to extract linkage from such a dynamic
web page.For example:
read link from Table mydb in database
<a href='<%=link%> '>No link in origin web page</a>



carantion (康乃馨) 于Sun Sep  8 20:54:37 2002提到：

在web使用挖掘中

数据预处理阶段的用户识别时，一般的要结合网站的topo结构来确定

如果动态网站中没有这种关系，用户识别的结果将比较的不准确

请问动态网站中，这是如何实现的？



sinokdd (KDD in China) 于Mon Sep  9 00:54:18 2002)
提到：


As I know, wget can be used to download a web site, thus it can
be modified to extract web topology, but for the dynamic page,
it doesnot work very well. 
I am now using MSHTML to parse the web page, and not in design
mode, thus it will load in web page just as IE, so if you can 
click a hyperlink in IE, you can get the hyperlink.

helloboy (hello) 于Mon Sep  9 18:40:08 2002提到：

How can get a runtime web-page's content?
Even if you can ,how can you deal with the problem that the links are created
dynamically. That is, the link are different according to the different context.


sinokdd (KDD in China) 于Mon Sep  9 23:37:49 2002)
提到：


No matter how these pages are, they will be shown in IE finally, and the
hyoerlinks will be steady while they are in IE, cannot change from time
to time, if so, no one would like to browse.
I am using MSHTML to load the html file, and call function in MSHTML to
parse it, just as what IE suppose to do, and get all the content of 
that page. You can image that the page is now displaying in the memory,
and by calling some functions from MSHTML you can access any content of
the page.

helloboy (hello) 于Tue Sep 10 08:22:50 2002提到：

I don't mean that.
For example,
<a href='<%=userid%>.asp'>userid连接</a>
linkage is dynamically link to [userid].asp according to
your input of userid. That is ,I will show personalization web page
to you when you log in. How can you achieve every user's linkage?