sourcemash / Sourcemash

Read Your News Faster.
GNU General Public License v2.0
3 stars 0 forks source link

Multi-Document Summarization #113

Open swglad opened 9 years ago

swglad commented 9 years ago

"This is usually called "multi-document summarization" in the literature if you're looking for a search term. Towards Multidocument Summarization by Reformulation: Progress and Prospects (http://people.csail.mit.edu/regina/my_papers/stim2.ps.gz) is one of the classic papers on this, so start there. Also look at http://www.cis.upenn.edu/~nenkova/1500000015-Nenkova.pdf for an overview of summarization techniques (including multi-document).

There were a series of workshops on multi-document understanding (http://duc.nist.gov/pubs.html). I recommend trying to implement ideas from the AAAI '99 paper first."

~ Sravana Reddy

swglad commented 9 years ago

Will be applied once for each category, and a summarized description of all articles in the category will be produced and rendered to the category-view.

alexgerstein commented 9 years ago

Maybe try sentence intersection to see which sentences are most important across articles as a first go?