beeseek-devs team mailing list archive

["Simone Brunozzi" <simone.brunozzi@xxxxxxxxxxx>]

Hi there!
First, I can't access the private section of our mailing list, because
I don't have a password for my email address. Can someone of you
provide it, or tell me where to set it up?

Back to the main topic, that is: "Relevant Factors".

Just to agree on terminology, let's say that:
Focus time: the time when the user is viewing the page and the windows
is not minimized or hided.
Scrolling time: how long the user scrolls the page.
Scrolling actions: number of times that user scrolls the page.
followed links in the page: links that are followed during the navigation.
Number of results viewed.
Page approved: if the user "approves" the page (if and when a
mechanism to catch that will be implemented).
Last search query: the query used to find the search results used.

I think that we may "start" with this algorithm:

values to store
- IP address, OS, browser
- search query
- results provided by google (pages)
- time before any result is clicked (longer time means the results are
not easy to read, or are not expected)
- which result is clicked, and in what order (1,4 means the first
result to be clicked is the 4th).
- how much time I spend on the clicked page (longer is better;
non-clicked pages have a time of 0 of course)

This could be a simple formula to rank pages for a particular query
(or a query that contains this query, such as "ubuntu" is contained in
"ubuntu drivers"):
rank of the page in google = A (from 1 to infinite, 1 is better)
time before any result is clicked = B (in centi-seconds, less is better)
order in which I click the result = C (from 1 to infinite, 1 is better)
time spent on the clicked page = D (in centi-seconds, more is better)

BeeSeek value of a page (less is better) =
500A + B + 800C - 1/D

Let me know what you think about it.

Simone


On 10/2/07, Andrea Corbellini <andrea.corbellini@xxxxxxxxxxx> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Andrea Colangelo wrote:
> > Regarding focus time, we should get informations about user activity. If
> > the user choose a page and get a phone call, the page stays there for
> > minutes and we would consider it important. We should find a way to
> > catch mouse move or keybard usage (like Skype or pidgin do to
> > auto-enable away mode).
> Great idea! I searched and I found that it is possible! Also we can find
> where the mouse is located, but I have no idea how can it help us.
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.6 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
> iD8DBQFHAkwnf3mN/RoI6JkRAtQdAJ43zmknqNYyV3cczFw5ZhY0wfX8UwCeMkCv
> D3hgKE2KRE7OG8FhTCU/ft8=
> =7BIl
> -----END PGP SIGNATURE-----
> _______________________________________________
> Beeseek-devs mailing list
> Beeseek-devs@xxxxxxxxxxxxxxxxx
> http://lists.beeseek.org/listinfo.cgi/beeseek-devs-beeseek.org
>


-- 
Ing. Simone Brunozzi

Via del Volontariato, 22 - 06083 Bastia Umbra (PG) - ITALY

Cell. +39 392-1551977 / +39 340-5768488
---------------------------------------
www.ubuntista.it | www.nonovvio.it
---------------------------------------