News for hackers. Stuff that matters.
Re:Ask HN: Extracting book titles from comments by garysieling on Wednesday April 26, 2017 @10:38AM (#14203327)
I've found experimentally that doing a search on amazon for a book title, even if you're only close, almost always turns up that exact book. However, the product API has really restrictive terms.

I would look at the OpenLibrary dataset - you might be able to match titles in there to comments, or use it to validate the NER output, if you don't want to go the Amazon link regex route.

The entire dataset is available for download, or you can build a prototype with their API - I did this to map speakers to books with https://www.findlectures.com (you can see it if you hover over a name - e.g. https://www.findlectures.com/?p=1&speaker=-Barack%20Obama).

There are no comments on this comment yet.