Traumwind - Some thought on similarity

As my readers have gatherd, I'm rather fond of my 'similar entries' feature I have running here. Now I often combine it with a quick keyword search using my brute force regex search.

One thought that allways shows up with any kind of 'autogenerated categorisation' is this: 'Why are those two entries related, or why are they considered related?'
Problem is, there are usually serveral kinds of relation entries can share. Same keywords, linking the same URL, by the same author etc.

I my experience thought rhat usually matters very little. Once one learns to take the results with some curiousness, and is expecting a surprise here and there, most relations auto-generated make perfect sense.

Given the realm I use this kind of FOA in, the 'surprises' are actually welcome. Boosting assosiative recall, finding buried knowledge, making new, unespected connections between materials.

[ by Martin>] [permalink] [similar entries]

similar entries (vs):

similar entries now working (# 11%)
keyword similarity works (too) (# 10%)
finally, site search (# 10%)
first results (# 10%)

similar entries (cg):

Brute Force Searchengine in PHP (# 11%)
simon (# 9%)
PERL OF WISDOM (# 8%)
similar entries now working (# 7%)
well well (# 7%)
finally, site search (# 7%)
keyword similarity works (too) (# 7%)

relevant words

Big things to come (TM) 30th Dez 2002

Into the impossible
Oblique Strategies, Ed.3 Brian Eno and Peter Schmidt

amazon.de Wunschliste

usefull links:
Google Graph browser
Traumwind 6-Colormatch
UAV News

finding (2.2)
usually (2.1)
same (2)
thought (2)
entries (2)
related (1.9)
are (1.7)
unespected (1.7)
fond (1.7)
surprises (1.7)
assosiative (1.7)
materials (1.7)
buried (1.7)
matters (1.7)
realm (1.7)
categorisation (1.7)
autogenerated (1.7)
aboutfoa (1.7)
combine (1.7)
gatherd (1.7)
auto-generated (1.7)
rhat (1.7)
curiousness (1.7)
recall (1.7)
boosting (1.7)
kind (1.6)
search (1.5)
serveral (1.5)
surprise (1.5)
brute (1.5)
learns (1.5)
relation (1.5)
my (1.4)
relations (1.4)
entries (1.4)
perfect (1.4)
regex (1.4)
expecting (1.4)
connections (1.4)
similar (1.4)
linking (1.4)
considered (1.4)
allways (1.4)
here (1.3)
there (1.3)
similarity (1.3)
why (1.3)
force (1.3)
welcome (1.3)
author (1.3)
sense (1.3)
kinds (1.3)
keywords (1.2)
shows (1.2)
with (1.2)
knowledge (1.2)
making (1.2)
quick (1.2)
i (1.2)
experience (1.2)
share (1.2)
some (1.2)
url (1.2)
feature (1.2)
readers (1.2)
of (1.1)
running (1.1)
keyword (1.1)
the (1.1)
take (1.1)
results (1.1)
often (1.1)
given (1.1)
one (1.1)
have (1.1)
once (1.1)
between (1)
is (1)
two (1)
etc (1)
make (1)
this (0.9)
little (0.9)
rather (0.9)
most (0.9)
they (0.9)
using (0.9)
why (0.9)
actually (0.9)
up (0.8)
those (0.8)
out (0.8)
any (0.8)
new (0.8)
use (0.8)
and (0.8)
a (0.7)
im (0.7)
very (0.7)
by (0.7)
now (0.7)
or (0.7)
can (0.7)
as (0.6)
on (0.5)
that (0.5)
it (0.5)
in (0.5)
to (0.4)
martin (0.3)
wisdom (0)
www.linux-mag.com (0)
/cgi-bin/printer.pl (0)
dissonance.no-ip.com (0)
dissonance (0)
incrementally (0)
userland (0)
users (0)
(0)
files (0)
perl (0)
(0)