W3C home > Mailing lists > Public > www-talk@w3.org > July to August 1995

Re: Accurate user-based log file analysis

From: Terry Myerson <tmyerson@iserver.interse.com>
Date: Tue, 18 Jul 1995 13:48:57 -0700
Message-Id: <199507182048.NAA05791@iserver.interse.com>
To: Brian Behlendorf <brian@organic.com>
Cc: www-talk@w3.org

Re: Interse' market focus

>> - The request media mix 
>> - The requests relative to main entry pages (ie the home page)

Brian-

There is actually some meat here. My hesitancy to elaborate is to
not give up competititve advantage. But of course I need to tread
lightly, in that we need to achieve some semblance of technical
credibility.

There are 2 big issues:

- caching

This is the easier one to solve. Caching can be approximated based upon the
media mix being requested from the site. Caching theoretically keeps the
request load constant at the server, when it peaks at the client's end.
However, the media mix being requested from the server does change. Of course,
the algorithm does need to identify special cases such as the Online services.

- gateways

This is indeed more difficult to solve. The requests must be de-interlaced.
If you know the structure of the site, you can make effective session
assignments to individual requests. I would not use this to plot the user paths.
I am very comfortable using it for aggregate statistics such as # of users,
session length, etc...

What are you guys implementing?

-Terry

-----------------------------------
Terry Myerson
Interse' Corporation
408 732-0932 x-230
408 732-7038 fax
tmyerson@interse.com
http://www.interse.com
-----------------------------------
Received on Wednesday, 19 July 1995 03:05:36 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 27 October 2010 18:14:17 GMT