[diamondap]

The ability to read the opening pages of the book is really cool. The results were pretty good for the searches I did, like a search for Stephen King's Mr. Mercedes bringing up a number of Dean Koontz books. It would be nice to have a little blurb with each title, so we don't have to judge by the cover alone.

Did you have to create the previewer yourself, or does that come with the Google books API? And how do you determine which titles are similar to a specified title?

[codemonky]

Thanks! I created the previewer myself in ReactJS/Ruby on Rails. The titles that appear in the search results are driven by the google books api.

[diamondap]

Well, it's a cool project overall. Definitely something you can fiddle with. Finding and recommending similar books seems to be a difficult problem. Amazon does a decent job, sometimes, though they tend to push top-sellers, making other titles invisible.

Goodreads does an OK job, some of the time, though most of the books I discover there come from reading my friends' good reviews.

RiffleBooks once looked promising, but they seem to have abandoned their site. It's been down for weeks. Like Goodreads, the main path to discover was following the reviews and recommendation of other readers whose tastes were similar to yours.

TheStoryGraph.com takes a unique approach. Readers tag titles with a handful of attributes, such as mood, pace, length, etc. If you like a book, the site can recommend others that have a similar mood, pace, length, etc. BookBub uses a similar tag-based approach.

If your site starts to pick up subscribers, you may want to consider the tag-based approach. It makes it easy for readers to contribute curatorial knowledge. It may also combine well with Google's recommendations.

That may be more than you want to think about for a hobby project, but I wanted to contribute my two cents.