top | item 4500323

Ask HN: Is there a future for automatic summarization startups?

1 points| MojoJolo | 13 years ago | reply

I'm currently doing my MS thesis which is about automatic summarization. I'm also creating a web app (http://www.indiegogo.com/TextTeaser?a=1110234) to showcase it.

Some startups doing automatic summarization are Cruxbot (http://www.cruxbot.com/) and Summly. Like those apps, my thesis is also doing extraction only. It just extracts the most important sentences in an article that is determined by several features.

So, is there a future for the apps (including mine) that I mentioned? Or is there a future for automatic summarization in general?

4 comments

[+] hbbio|13 years ago|reply

Depends. I was familiar with http://www.informatik.uni-trier.de/~ley/db/indices/a-tree/a/... and to be successful, it has to be a great product. I.e. find where the pain is and solve it. Success lies on the product side, not the research...

[+] MojoJolo|13 years ago|reply

Turning the research into a product is the problem. But I admire what Lytro has done. His dissertation turned into a real product.

Hmmm. But what do you think about those apps that I mentioned? They have pretty good algorithm inside, but are they great products?

[+] scotto|13 years ago|reply

I believe there is, but like hbbio said it's a matter of the product. You can make the algorithm into a product by offering it through an API to other developers.

[+] MojoJolo|13 years ago|reply

Yes, APIs can change an algorithm into a product. This was done by Diffbot (http://www.diffbot.com/) which presents different text algorithms. I love their API, and currently using it for my thesis.