What is Percevia? |
|
Basically Percevia is a web
based search
engine designed specifically
for quickly finding objects
in any collection. Percevia
works with any thing that can be
organized into a collection,
provided they share common
characteristics. |
| |
So how is
Percevia different from a
search engines like Google?
|
| Google
scans the web and gives you back
links to web sites that may or may
not be
relevant to the text you entered.
With Percevia you don't type
any text, and it does not
search the web, rather with
Percevia you click on icons
and it scans its built in
database. Furthermore
Percevia
does a step by step search,
and at each step it narrows
down the matches in the
database. |
| |
Why is that
better than simply typing
the name of your bird into
Google? |
| There
are two major reasons. First,
in order to find what you
are looking for on Google,
be it a bird or a laptop or
a fact about the world,
Google gives you back
"hits" which are
links to web sites that
contain the text you entered.
So you have to visit each
site to see what it offers.
You have no way of knowing
what you will find. Second,
Google does not help you do
any kind of identification,
for example if you saw a
bird with a crested plume
that was black, entering
"crested plume black bird"
in Google just gives you
thousands of sites that you
must examine one by one.
Percevia on the other hand
will find every bird in its
database with a black
crested plume. You can see
what they look like in a
list of icons, and you can
click on any of these icons to get
more information. |
| |
Okay I think
I understand the difference,
but don't I still do
a lot of clicking in
Percevia? |
| Not
really. Almost any bird can
be identified in just a few
clicks. Especially if you
have been observant and
noticed a distinguishing
feature. There are over 50
different attributes in the
bird database that you can
search on, and there are
over 700 different birds.
Finding a bird usually takes
a few minutes at most. |
| |
Is Percevia
a commercial product? |
| It was.
We once sold a desktop
version of Percevia,
thinking people would want
to use it to organize there
own stuff, build collections
to search in clever
ways. But we eventually
discovered that was not a
large market, people don't
have time to build
databases. |
| |
Do you offer
some kind of service to make
money? |
| Right
now we are offering the bird
search engine for free. We
built it originally as a
demo of the underlying
technology. But because I
love birds so much I wanted
it to be the highest
quality. That turned out to
be a real challenge, for
example you need great
illustrations, bird call
audio examples,
extensive species
information, and so on. I eventually
acquired all that and we
suddenly had thousands of
people visiting the site. |
| |
Will
Percevia ever be a
commercial product? |
|
Eventually we want to market
it to webmasters to give
them a better way for people
to search there data. But
that's a ways off. |
| |
So how do
you support all this work, I
mean it must be expensive
with all those great
drawings? |
| We have
some Google and Amazon ads
running but they hardly pay
the bills. At some
point we might make this a
subscription-based site and
charge some minimal fee to
have access to all the data
but right now
it just feels good to know
that I am helping people who
love birds find them and
learn about them. |
| |
Ok so much
for altruism, can you tell
me more about why people
would use Percevia over
others that do the same
thing? |
| Well it
turns out there really are
no other sites that do the
same thing as Percevia,
which is identify birds in a
step by step manner. What I
mean by this is that while
there are web sites that
claim to help you identify
birds, they don't really do
that. Let me give an
example. One of the larger
sites is called
The Bio Diversity Institute
Internet Field Guide to
Birds. You can click
that link to see it. What it
does is typical of most of
these sites: you are
presented with large
collection of field mark menus. You
select attributes of the
bird from these such as
geographic area, body size,
shape, wing characters and
so on. After selecting all you
know about the bird you
click a Submit button and
the site
finds every bird that
matches your choices. Sounds
good but what you
get back are either dozens
of birds or, if you picked
characteristics that don't
match any bird, you get back
nothing. Not a great
experience. |
| |
And how is
Percevia different? |
| With
Percevia you select just ONE
characteristic of the bird at a
time and that makes all the
difference. For example you can
pick something called
Location. You are presented
with all the states,
provinces, coasts and
islands in North America.
You select the one where you
saw the bird, say
California, and then the
search engine presents you
with a list of all the birds
that are in California. We call
these the matches. Now
you pick a second attribute,
say Size. As you go though
this process the number of
birds in North America is
reduced until you find your
bird. |
| |
Aren't there
other examples on the net
that help you find a bird? |
| There
are other sites that help,
but they do it in a crude
way. I am not disparaging
these sites, just pointing
out that while they might be
informative, they are not
good at identifying. Take
enature.com, one of the
most popular destination sites for
learning about nature on the web. They
present a simple list of
bird shapes. We have that
too, but the problem with
enature.com is that the
search ends at that point:
you click a shape and are
presented with a matrix of
tiny photos of the birds with
that shape. You have to
click on each one to see a
larger picture and figure
out if it looks like your
bird. I would call this akin
to a regular book field
guide, not an ID engine. |
| |
Anything
else you want to tell me
about Percevia that makes it
special? |
| Perhaps
one of the most interesting
aspects of Percevia is one
that the least obvious,
which is something we call
smart attribute and value
elimination. When you do
a search, Percevia keeps
track of each attribute and
value you select. After each
step the attributes and
values are filtered to just
those that are relevant to
the remaining matches. So
for example if you have
picked a set of birds that
happened to all be the same
color, there is nothing
gained from presenting color
as a selection, and so that
attribute is eliminated from
the search window. Same
thing with values. If those
matched birds do have
different colors but none of
them have black as a color,
the value "black" is not
presented in the color
attribute. This might not
seem very profound, but it
seriously reduces the
possibility of picking a
value or attribute that will
lead to zero results or not
narrowing the search. This
is why we say "always a
result" because with smart
attribute and value
elimination there is no way
you will end up with no
birds found. |
