|
This is our first quarter since remodeling the Physics web site.
I've adjusted the analysis accordingly; our file names and directory
structure might have changed, but the content tracked here is all
consistent. I have also started tracking all of the "top-level" Physics
menu pages, though they don't show up here since there's only one
quarter of data so far.
In late May 2006, I was able to subscribe the JRM site to
Google Analytics,
a free service that provides a variety of traffic analysis tools that
leverage Google's enormous infrastructure.
A new page has been added
with screenshots of some of the more interesting of that data. Google
Sitemaps, now renamed
Google Webmaster Tools,
also provides useful data which is included with the above.
|
|
|
Total traffic to all
Physics web sites as a function of time.
The traffic
is shown as both the average pages served per day (green) and as the
total number of distinct hosts served per quarter (blue). Exponential fits
to the data are shown.
It's a bit curious to have the "page views" and the "hosts served" trending in
opposite directions, but that's probably an artifact of the search engines and
confused visitors having to figure out our new site organization.
|
|
| |
|
Here's just the traffic to the departmental parts of the main
Physics server (including the personal sites) plotted with the
old all-of-Physics traffic data. This isn't fair, of course, since
many of our hits are now on separate virtual servers, but
with the above it helps to give some idea of how things trend.
The fits are quadratic.
|
|
| |
|
These are the total hits per quarter on the most popular Physics web pages
as a function of time.
Only research and teaching pages are included.
Historically, other utility-type pages, such as the personnel directory,
have been omitted due to difficulty tracking them in our previously haphazard
site structure. The new site is cleaner and those pages are being tracked now.
I've added the HEP,
Origins, and
PCSC sites. Recall that
HEP has other computing assets that are not tracked here.
The PCSC has just remodeled their page and given it its own site name,
so watch for it to grow. The spike in MSDS hits in the first quarter 2006 is due
to an overeager robot.
|
|
| |
|
This is the same data, but with the top two pages (the Physics home page
and the JRM MSDS pages) suppressed. This makes the other pages easier to
evaluate. Overall our pages seem to be approaching a stable level of usage.
Note how class-oriented pages show an oscillation with semesters not present
in the research-oriented pages. This is the sort of page now hosted at
K-State Online.
|
|
| |
|
Total traffic to the
JRM web site as a function of time. The traffic
is shown as both the average pages served per day and as the total number
of distinct hosts served per quarter.
The solid line is a linear fit to the page data. The data still reasonably match
the linear fit, despite the occasional spikes.
|
|
| |
|
The most popular pages for the JRM Lab, minus the home page and the
MSDS pages. The robot-induced spike in the gallery and policy pages
has subsided, but the laser safety pages are off the chart. That seems
to be from someone linking to our
eye safety horror stories reprints.
|
|
| |
|
The popularity of the galleries and tutorials obscures the trends in the
other JRM pages in the graph above. This graph suppresses those
pages. Newly added is data for our laser safety pages, as noted above.
By and large growth is steady for most pages outside the safety and
computing categories.The demand for that information is evergreen.
|
|
| |
|
This is a comparison of the JRM web pages viewed
by internal, KSU users versus those viewed by external visitors.
Each category of pages is shown by its percentage of the total
page views for that kind of visitor. Note that the internal
home page views and the external MSDS page views are both off-scale.
If you don't like the pretty 3D view at right, there's a
2D version.
As you might expect, the KLS
schedule,
personnel pages and
AMO seminar
schedules are more interesting to local users. You may also see
the raw analysis of just the local in-house traffic for
Q2 06,
Q1 05 and for
Q4 05.
|
|
| |
|
My old "Vince's Server" now no longer exists; this data is
actually a synthesis of the traffic to my
personal pages, which
have been reorganized on the new JRM virtual server.
The lines are third-power fits to each of the data sets. Again,
the human-driven unique hosts grew while the robot-skewed page views
stabilized. The generally strong growth is principally due to my genealogy
pages.
|
|
| |
|
Traffic for the Physics Education Research Group
(PERG).
The fits are 3rd-order. Again, unique hosts have increased while
page views are stable. Generally speaking,
traffic to PERG is broadly spread over the
VQM graphics and pages.
|
|
| |
|
PERG also has a new initiative called the
Physics Pathway
(Physics Teaching Web Advisory). The use of this
rather special-purpose site is obviously episodic.
The fits are just to guide the eye.
|
|