Ceph is SUCH a geek...

[Cepheus 0]

http://68.58.244.113/rng/index.asp

And that proves it...

That's a Random Name Generator, programmed by yours truly. Pain in the ass took me 2 days straight coding to write, but it is, if I do say so myself, one of the best generators in existance... (I'm not prone to bragging, so this is a big step for me! lol)

Instead of using hard coded letter combinations and crap like that (like most of these generators), it accepts a list of words (most any languages), then splits the words into their syllables, and collects statistical data about how they connect to other syllables. All this is done behind the scenes -- the result is that every single word is pronouncable (as long as the input was -- garbage in, garbage out).

While most of these generators produce maybe 2 or 3 usable, pronouncable words in 10 (if you're lucky), mine has been capable of producing decent words in nearly all its runs. Of course, tastes vary...

Additionally, my generator has a feature called Extrapolation. Feed it a word, and it'll break it down into its syllables. Select which syllables you want it to change, and it'll show you a list of words, extrapolated based on the data you've given it.

What does all this mean?

I haven't jumped in 3 weeks...... Ceph has WAY too much time on his hands.

--
Skydive -- testing gravity, one jump at a time.

[AirMail 0]

You win the Geekier Than Thou Award.

I just found my new password generator. Nobody will be able to guess these.

Patrick

--
It's never too late to have a happy childhood.
Postal Rodriguez, Muff 3342

[Cepheus 0]

Ooh -- hadn't thought of that! Excellent!

--
Skydive -- testing gravity, one jump at a time.

[DZBone 0]

Nice. What kind of stats did you calculate?

When we were trying to think of a name for my soon-to-be-born son, I wrote something similar, but at the letter level. It chewed on a list of names, and generated made-up names using the probability of next-letter occurance. It was awhile ago, but I think you could give it a starting letter and word length, and it would generate the "most probable" sequence of letters given the corpus.

Needless to say, the wife vetoed the names it came up with...

_________________________________________________
If you hadn't read this, would it have made a sound?

[Cepheus 0]

Actually mine does something very similar, only on a syllable basis. It splits every word into their syllables, the counts how many occurances of each type they are, and pops em into the database. It also analyzes which letters 'connect' the two syllables, counts them, and adds them as Letter Pairs into the database (like in the word 'generate' which splits into 'gen|er|ate', 'NE' and RA' are two letter pairs.

The generater uses this frequency data to first select a starting syllable, then it asks the database for all the Letter Pairs that start with the last letter of that first syllable. It them picks a random letter pair, then asks the database for all the syllables that start with the last letter of the Letter Pair. The process continues, splicing syllables until the desired number of syllables creates the word.

I actually wrote the Syllable Stat Collection program about 3/4ths of a year ago, while sitting around bored... I slapped a simple interface on it, and that's what I use for the back-end data collection. The generator itself I wrote a few months later, and has been sitting like that until today, when I finally created the Extrapolation feature I've been itching to do.

Extrapolation basically does the same thing, only it has to make sure the syllables that are getting replaced match up with the existing syllables. Wasn't too hard -- took me bout 2 or 3 hours. I've been wanting to do this for a while, but I never had the motivation. I thank AlliKitten for that.

--
Skydive -- testing gravity, one jump at a time.

[allikitten 0]

Glad I was able to help. **hugs**

*whispers* But I have to admit, I had an ulterior motive. This is a great name generator....



Great work.

>^..^Allikitten