Amazon also has a massive database of associations between books. Richard Herley presents an extended example based on associational results, so I thought some readers might be interested in the type of data currently available through Amazon for the example book 1984. Amazon’s associations are based on search and purchase history and not on “a list of favorite books”:

Customers who searched for “1984″ also expressed interest in:

Animal Farm and 1984 by George Orwell (2003)
Brave New World by Aldous Huxley (2006)
1984 (Signet Classics) by George Orwell (1961)

Customers Who Bought This Item Also Bought

Fahrenheit 451 by Ray Bradbury
Brave New World by Aldous Huxley
Catch-22 by Joseph Heller
The Catcher in the Rye by J.D. Salinger
Lord of the Flies (Penguin Great Books by William Golding
Slaughterhouse-Five by Kurt Vonnegut
Animal Farm (Signet Classics) by George Orwell
A Clockwork Orange by Anthony Burgess
Brave New World and Brave New World Revisited Aldous Huxley
The Great Gatsby by F. Scott Fitzgerald
Of Mice and Men (Centennial Edition) by John Steinbeck
To Kill a Mockingbird by Harper Lee
One Flew Over the Cuckoo’s Nest by Ken Kesey
Frankenstein (Signet Classics) by Mary Shelley
Frankenstein (Enriched Classics) by Mary Shelley
The Stranger by Albert Camus
The Grapes of Wrath (Centennial Edition) John Steinbeck
1984 (Cliffs Notes) by Nikki Moustaki

These titles are from only the first 3 of 17 “pages” of books. The books overlap with Herley’s envisioned results; however, The Handmaid’s Tale by Margaret Atwood does not appear until page 5.

These initial titles show a problem with Amazon data I think. It is probably distorted by book lists in schools that define “required reading” and “suggested reading” in classes. For example, the connection between 1984 and The Great Gatsby is not based on theme, genre or style. Results based on “favorite books” would be interesting to see.

Kevin, sorry, there’s nothing to report.

Here’s a description from their site:

BookLamp.org matches readers to books through an analysis of writing styles, similar to the way that Pandora.com matches music lovers to new music. Do you like Stephen King’s It, but thought it was too long? BookLamp allows you to find books with a similar level of tone, tense, perspective, action, description, and dialog – while at the same time allowing you to specify details like… half the length. It’s impervious to outside influences – like advertising – that impact socially driven recommendation systems, and isn’t reliant on a large user base to work.

P.S> I am not associated with Booklamp in any way. i just think they have a cool idea and might come out with something good eventually

I’ve spent (not enough time during) the last year running a fairly large community site on a Drupal backend (based in part on a similar theory of having highly ranked content float), and would love to chat with you more about it if you’d like.

I agree that niche-related reviews might drive interest in the beginning, but also think that in this case starting big is a good idea. Perhaps there could be channels where reviewers who specialize in a particular genre could be highlighted; this might also be a great way to get genre authors involved by highlighting their work through interviews, Q & A, etc.

(This is my first time reading your blog; I’m looking forward to catching up!)

On starting a (e)book review site: it is easier to get established in a niche than it is with a generalist site. Because of this I think a genre specific site might catch on (e.g.: Mystery & Crime, Science Fiction and Fantasy, etc.) because you are more likely to attract readers and reviewers who have a passion for that subject.

Likewise, I think it is preferable to have the site be ebook-only, again as part of the “own the niche” strategy. Filters are fine, but the search engines will tell you most people do not use filters or drop down boxes or radio buttons when they use a search engine, and they are unlikely to use those on a site search too.

I think there is currently a large gap between the parties involved in books (regardless of format)…so far I think services seem to be focused on tying people of the same group together (ie. readers to readers, pubs to pubs, reviewers to reviewers)…to make things worse they are not really doing any of it in ‘ingenious’ ways that make people talk about it (librarything is probably pushing the envelope the most in this world right now)…

What I think needs to be figured out is a better system that ties readers to reviewers to authors to publishers (and to sales data)…it’s a little bit Amazon, a little bit facebook, a little bit library thing, a little bit lulu, and a little bit J’ne sais quoi…

The first challenge is collecting the data (the right data), the second challenge is figuring out the right filters and applying the right glue…that’s just my two cents.

Kevin, that sounds great! We’re still discussing this at the Teleblog, but I’ll mail you soon with a résumé of where we’re at so far.

I hope to make it public very soon.

For big sites with database connections, librarything is fulfilling that purpose. Otherwise, the best consumer reviews are on ebook marketplace sites. Literary bloggers have not seen a particular need to focus specifically on ebooks (because many writers are still coming out with print books first).

One of the problems with review sites is that they are either too highbrow (commenting on only a small number of titles) or too eclectic (mixing celebrity bios with home repair books with fiction). Even in the realm of creative writing, there is a lot of ground to cover (and to set up categories for).

Two sites I go to for reviews: Scott Esposito’s Quarterly Conversations and M.L. Orthofer’s Complete Review . I’m not too familiar with speculative fiction, but there are several excellent review sites. Also, I’m finding a lot of good recommendations from mobileread’s forums, although forums are hardly the ideal format to have a literary dialogue.

Perhaps we can set up a wiki or something and start fleshing more of this idea out and I’ll get to work on a prototype of some sort…