Announcing Goodreads Personalized Recommendations
Goodreads was founded with the mission to get people excited about reading. And the key to getting people excited about reading is to help them discover books that they’ll love, and then to enable them to share their thoughts and experiences with friends.
Today, Goodreads launches a new personalized book recommendation engine. It takes recommendations to a new level of sophistication by analyzing both books and, more importantly, the people who read them. It’s the Netflix of book recommendations.
Finding a great book recommendation online has been a hit-and-miss affair to date. We’ve all experienced the unhelpful suggestion to read another book by an author we already love. And how about the dreaded impact of buying gifts on Amazon only to have irrelevant book recommendations come up for months afterwards?
Earlier this year, Goodreads purchased a company that had built a very sophisticated book recommendation system. Today, after months of hard work, we’re ready to provide you with book recommendations that take into account what you like and don’t like and what certain books mean to you.
To get started, rate at least 20 books (and rate much more to get even better recommendations). Categorize your books in custom shelves that reflect what the books mean to you. Then explore your recommendations. We apologize for the impact this will have on the size of your to-read shelf.
How Goodreads Recommendations Work
The Goodreads Recommendation Engine combines multiple proprietary algorithms which analyze 20 billion data points to better predict which books people will want to read next. It maps out the connections between books by looking at how often they appear on the same bookshelves and whether they were enjoyed by the same people. On average, Goodreads members have 140 books on their shelves. With this information, the engine learns how your tastes are similar to or different from the tastes of other Goodreads members.
So, a big part of the secret sauce is…you, the Goodreads community. The Goodreads community is almost six million members strong, and you’ve added a combined total of 190 million books to your shelves!
Take best-seller The Help as an example: Goodreads members have added over 175,000 ratings of the book and over 40,000 reviews. In comparison, Amazon has, to date, less than 4,500 reviews and ratings.
But it doesn’t end with raw numbers. Goodreads members have put The Help on bookshelves called Historical Fiction, Friendship, Racism, Women’s Fiction and Cultural > African American. You can see how this book means different things to different people. If you’re approaching this as Historical Fiction and have enjoyed The Guernsey Literary and Potato Peel Society and A Tree Grows in Brooklyn, then a great recommendation for you is These Is My Words. With Amazon, the focus is on other best-sellers so someone buying The Help would get recommendations for books as diverse as Water For Elephants, The Hunger Games, and One Day.
We welcome you to try our recommendations on for size. Compare them with anything else you’ve relied on online. Then tell us (and your friends) what you think. We think you’ll be blown away, and that you’ll meet your next favorite book on Goodreads.
Today, Goodreads launches a new personalized book recommendation engine. It takes recommendations to a new level of sophistication by analyzing both books and, more importantly, the people who read them. It’s the Netflix of book recommendations.
Finding a great book recommendation online has been a hit-and-miss affair to date. We’ve all experienced the unhelpful suggestion to read another book by an author we already love. And how about the dreaded impact of buying gifts on Amazon only to have irrelevant book recommendations come up for months afterwards?
Earlier this year, Goodreads purchased a company that had built a very sophisticated book recommendation system. Today, after months of hard work, we’re ready to provide you with book recommendations that take into account what you like and don’t like and what certain books mean to you.
To get started, rate at least 20 books (and rate much more to get even better recommendations). Categorize your books in custom shelves that reflect what the books mean to you. Then explore your recommendations. We apologize for the impact this will have on the size of your to-read shelf.
How Goodreads Recommendations Work
The Goodreads Recommendation Engine combines multiple proprietary algorithms which analyze 20 billion data points to better predict which books people will want to read next. It maps out the connections between books by looking at how often they appear on the same bookshelves and whether they were enjoyed by the same people. On average, Goodreads members have 140 books on their shelves. With this information, the engine learns how your tastes are similar to or different from the tastes of other Goodreads members.
So, a big part of the secret sauce is…you, the Goodreads community. The Goodreads community is almost six million members strong, and you’ve added a combined total of 190 million books to your shelves!
Take best-seller The Help as an example: Goodreads members have added over 175,000 ratings of the book and over 40,000 reviews. In comparison, Amazon has, to date, less than 4,500 reviews and ratings.
But it doesn’t end with raw numbers. Goodreads members have put The Help on bookshelves called Historical Fiction, Friendship, Racism, Women’s Fiction and Cultural > African American. You can see how this book means different things to different people. If you’re approaching this as Historical Fiction and have enjoyed The Guernsey Literary and Potato Peel Society and A Tree Grows in Brooklyn, then a great recommendation for you is These Is My Words. With Amazon, the focus is on other best-sellers so someone buying The Help would get recommendations for books as diverse as Water For Elephants, The Hunger Games, and One Day.

We welcome you to try our recommendations on for size. Compare them with anything else you’ve relied on online. Then tell us (and your friends) what you think. We think you’ll be blown away, and that you’ll meet your next favorite book on Goodreads.
Comments Showing 151-200 of 261 (261 new)
message 151:
by
Jan
(new)
Sep 16, 2011 02:48PM

flag


It is a really great idea and it looks good - although I am getting recommendations for books I have read too (?).
Well done you bad people :)


Mine is over 200 and its not complete.
In case anyone else is wondering…
I couldn't see my "read" shelf in the "Recommendations by Shelf" list, so I visited http://www.goodreads.com/recommendati...
I was shown a message saying "We are currently generating recommendations for your read shelf. Please check back soon."
A few minutes after visiting that page, I got recommendations from my read shelf. I'm hoping this is a beta thing :)
I couldn't see my "read" shelf in the "Recommendations by Shelf" list, so I visited http://www.goodreads.com/recommendati...
I was shown a message saying "We are currently generating recommendations for your read shelf. Please check back soon."
A few minutes after visiting that page, I got recommendations from my read shelf. I'm hoping this is a beta thing :)
The recommendations page is a wonderful idea! I'm already loving it.

BUT
I get an error each time I try to mark a book as not interested. Using Firefox 6.0.2 on Mac Lion, is it an issue on my side or on Goodreads?

BUT
I get an error each time I try to mark a book as not interested. Using Firefox 6.0.2 on Mac Lion, is it an issue on my side or on Goodreads?"
This is a known issue that we're working to fix. Sorry about that!
Edit: should be fixed now!




I hope that you will bring back "random" with the enhanced ability to randomize any one of our shelves. I would like to be able to randomize my huge "to read" shelf in order to select my next book. (I do it manually with random.org now, but it would be easier if it were integrated into Goodreads.)

You just need to click on the little arrow to say 'read' or 'to read' or another shelf. That should do the trick! :D



Known bug that affects a small percentage of users. Our team is looking into that too!
Edit: The bug has been fixed, so your recommendations should now be available.

Works fine on IE9.
On my phone, only Opera mini and Opera mobile seems to open the recommendations page. Ucbrowser, bolt, dolphin and the native browser all crash.
Why so?

- I consistently gave only one star to Dickens every book I saw (at least 4) yet a Dickens book is recommended for me! This gives the impression that the algorithm is taking note of the NUMBER of Classics I high-scored rather than the AUTHORS

- When I tried to mark Dickens rec as "not onterested" the page came up with an error message
- When I tried later to give a score to a rec which I have already read, another error message.
But it loooks fun! Certainly a quick way to give a score to lots of books.

Absolutely. You can create your own exclusive shelf for "did not finish" on the "edit bookshelves" page. Just "add a shelf" using the feature on the bottom of the page and check "exclusive" next to the new shelf when it appears on the list. As an exclusive shelf, it will be mutually exclusive with your "read", "to-read", and "currently-reading" shelves. Let me know if you have any trouble with this.

yikes.

Very. havent bought books since July to try and cut down on the tbr list.

I'm reading like a book a day, but they usually aren't from my list, because I can't go to my school library so I'm just reading what I can for the most part.

I have a list and I know I can get some from my library and these include e-books and audiobooks. You might want to check to see if you can get e&audio books without leaving home.

Great Job, Goodreads!

I can't take audiobooks. It just feels like I'm cheating.

I've refreshed several times and still no go.

I've refreshed several times and still no go."
i had the same problem. i used a different user name and it had recommendations as soon as i had 20 ratings. one of the employees said it's a known bug with some accounts.

This is AWESOME!!! Thank you Goodreads you've just added another reason as to why you are my favorite site on the web!

You can mark those you've read by rating them with the stars underneath "want to read". Or you can hover over the arrow next to "want to read" and select "read" to add that book to your "read" shelf. The algorithm will then know you've read that book and you don't have to mark "not interested".


Also, there is no way to mark as 'already read'. I haven't gone through and put on every book I have ever read in my whole life one-by-one; I am bound to run into things I have done already in your recommendations, but I can't indicate such.