14 Apr 2000 pate   » (Apprentice)

being here:

Hmm, mjs certified me as an apprentice, thanks. In attempt to prove that he's over-rated me, i thought i'd put some musings here. i don't think these are worthy of an article, but hopefully they're not too foolish ...

what i do:

well, i'm pretty pathetic as a coder. although i'm starting to do things that approach real work in perl. mostly, i'm a unix/network admin. i've been preaching the Free Software gospel for a while, and even seen it pay off a time or two.

my thoughts for today:

1) why does everyone who writes an intro to using sh derived shells (esp bash) avoid talking about functions? i got hit with another intro article yesterday, and in a fit of insanity put this together.

2) Within a community, it should be possible to build not just a trust metric , but an interest metric as well. I'm not talking about a content filter, rather a method for bringing people, articles, and perhaps external links regarding areas of personal interest to a higher level of visibility to the individual user. I'm not sure of the exact method, but would like to think aloud for a bit and see if I spark any ideas with the rest of you.

At the root, my thought is that several (perhaps 20) key terms could be identified, each having potentially several synonyms. Then every page a user visits, creates, or links to would be scored according to their use of those terms. A page being visited would be worth a low value to the reader for any terms it contains, a page that the user wrote would be worth more, and a page the user linked to would be worth still more (perhaps values of .01, .25, and .35 respectively). This attempts to reflect the increased interest in a field represented by the effort to write about it, or create a link to it.

Some mechanism for aging probably makes sense, maybe totaling a persons 'interest matrix' on a daily basis and determining their current interest with a moving average. Again, I'm not sure.

The net result of this is that each person would have an interest matrix, which could be used to weight objects for presentation to them when they've logged into the site. As a person sees more contributions in areas s/he is most interested in, they are likely to identify the people involved in creating them and responding to them ... hopefully encouraging more communication within the sub-group.

Good places to look for implementation ideas would probably include the news weighting system in Emacs (GNUS?), and the scoring mechanism used by ht:\dig or other search engines.

Latest blog entries     Older blog entries

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!