For my second pass at my Corporate Motto mashups, I recombined mottos by splitting each motto based on a certain number of words and recombining the mottos into a new motto.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 |
import sys import random stuff = list() for line in sys.stdin: line = line.strip() stuff.append(line) slogan_a = random.choice(stuff) slogan_b = random.choice(stuff) slogan_c = random.choice(stuff) #print slogan_a[:10] + slogan_b[10:] slogan_a_words = slogan_a.split(" ") slogan_b_words = slogan_b.split(" ") slogan_c_words = slogan_c.split(" ") first_part = " ".join(slogan_a_words[:4]) second_part = " ".join(slogan_b_words[3:]) third_part = " ".join(slogan_c_words[5:]) print first_part + " " + second_part + "." + " " + third_part #print first_part + " " + second_part |
The first round of mashups returned some funny results, but mainly mottos that almost sounded like they could be real mottos. Some of my favorite ones:
- Even your best friends won’t tell you!
- Snap! Crackle! Pop! just one!
- You’ve come a long way, baby.
- We Help You life.
- It looks good, it cleans your teeth.
I then took the first results, and passed them back into my corporate motto program. Round two returned much more ridiculous mashups:
- Finger-lickin’ good! you’re Zestfully billion!
- You never had as a baby. A drier bottom!
- Betcha can’t eat just where you live
- Plop, plop, fizz, fiz, it absolutely necessary to own two or more motor cars.
- Some girls have developed just one!
- nothin’ says lovin’ tell until I discovered Smirnoff someone
- Your World. Delivered. Touch someone needed.
- Reach out and touch tell you!
- Dryest gin in Future. love to touch it tastes good, and by golly it does you good
- Be a Pepper! clean friends won’t tell you!
At this point, I decided to combine 3 mottos instead of just two using the same initial first pass of mashups.
- It’s Your World. Take clean until you’re Zestfully clean all!!
- Finger-lickin’ good! discriminating family clean, it’s got to be Tide.
- It’s Your World. Take nicest people on it, flaunt it!. you!
- You deserve a baby . needed.
- When you say clean, baby something you never had as a baby. A drier bottom!. you never had as a baby. A drier bottom!
- I’m lovin’ It! 2 hands best teacher in choosing a cigarette. your teeth.
I really like the results, which have become so base-level in communicating intention that they confront you with the manipulative nature of the mottos.
I did this one more time, taking the second round of mottos and running them back through my program.
- break like it, flaunt it!. tell you!. you!
- Have love to touch. Have Depot, low prices are just the beginning.. what a relief it is!.
- Control. laxative starts to work? tell you!
- Dryest gin in Future. 2 Take Control. laxative starts to work?..best teacher in choosing a cigarette. your teeth.
- Be a Pepper! clean – billion! I discovered Smirnoff someone
- Some girls have developed just where you live. low prices are just the beginning..what a relief it is!
- Snap! Crackle! Pop! just 2 Take Control. laxative starts to work?.
- Some girls have developed touch tell you!. you live
- Be a Pepper! clean – billion!. I discovered Smirnoff someone
These last results make me think of a very drunk person trying to give sage advice, and failing miserably.
Further Exploration
I would like to write the program to return 100 results in a text file, and to automatically take each output and generate 100 new mashups using the previous output until a certain number of times (say 10). Each output would be saved in a separate text file. I would also like to randomize where the lines are being split between a certain range of numbers. I also noticed that some lines from the original text showed up frequently in the results, I assume this means I need to write a smarter random function.
I am also interested in ways I can analyze the structure of the mottos and recombine them at smarter moments. I think that would return results that sound more like a single sentence, and less like drunk ravings and/or bad English.
It might also be interesting to do some vowel translation on these new generations.
