MediaTakeOut Headline Generator
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
.gitignore
README
generated.txt
headlines.txt
headlinesprepped.txt
mto-analyze.py
mto-cfg-nltk.py
mto-cfg.py
mto-languagemodel.py
mto-scrape.py

README

MediaTakeOut Headline Generator from Robert Elwell

I love MediaTakeOut. So I decided to make a language model for it. 

I ain't copywriting shit, all this open-source software is legit. 
I don't mind if you look around the code a little bit.

mto-scrape.py 
	Run this first. It grabs headlines from MediaTakeOut. 
	It's really that easy. Takes <1hr.

mto-analyze.py
	Run this to get some statistics about what you pulled. 
	You need the sentences.prepped file it generates to run 
	your language model.

mto-languagemodel.py
	This actually builds out the fake headlines. Because I'm lazy 
	and I'm using NLTK's Text and NGram language model classes.
	I'm just writing to a file from stdout to get my sentences.
	Options are provided in this, so take a look.

generated.txt
	Here's an example of headlines. 
	Take a look at it in action at 
	http://robertelwell.info/mediatakeout-headline-generator/

Requires NLTK with the punkt and stopwords dictionaries downloaded.