Wednesday, November 12, 2008

The Google Registry

It would be a shame if the opportunity to create a unified (and uniform) bibliographic database itemizing the copyright status of our collective published works only resulted in a copyright database. Perhaps that comment appears strange but I believe the establishment of this 'registry' under the terms of the Google/AAP/AG settlement opens up a opportunity to tie an alphabet soup of identifiers, bibliographic (and multi-media product details), full content and transaction capability together in one form.

A primary reason ISBN was implemented so successfully in the publishing industry was that the number rapidly became integral to the publishing supply chain. Newer standard numbers such as ISTC, DOI - even ISSN in the retail channel - are not widely adopted because they have not become transactional identifyers. The same will be true if/when an Interested Party Identifier is established. The development of the Google Registry (surely the first task of the Executive Director is to come up with a name) may represent the only opportunity to tie each of these ancillary identifiers to one common objective. That objective is the better identification of content - not simply 'book' content, but content we will be using and transacting in an on-line environment.

How the registry may be formed is anyone's guess, but for sake of argument I envision a pyramidal structure. The identifier segment forms the pointy top layer, bibliographic data the second layer, content the third and the 'transaction gateway' the bottom tier. Then again maybe it's a cube and I should be adding subjects, a retail/library segmentation, and transactional details like rights information. Regardless, it seems to me combining each of these segments into a registry might engender significant opportunities to improve the publishing supply chain. But more than that, the combination I suggest works better for the on-line world than the off which is the failing of the current crop of ISBN databases (including

Merchandising and search would be vastly improved if we were able to search by ISTC or IP using one database that would return all renditions of a work or all works/items produced by an author. Additionally, access to the content would be immediately available providing views of the content the searcher was interested in. Of course, all the copyright details about that work would also be available. Importantly, each of the elements in this registry would be linked so if someone happened on work they could rapidly find associated versions (ISTC) or other content produced by the author or publisher (IP).

The most obvious application enabled via the 'transaction gateway' would be purchase but a 'transaction' can be many things: views, queries, checkin-out, use rights, syndication and may more. An open service architecture would enable development of third party API's that could result in all kinds of new applications but existing ones would also benefit as well. Worldcat and Copyright Clearinghouse applications are good examples where users could find the physical content in a library or attain usage rights from CCC.

Google has provided $35mm to fund this registry and the governing board including publishers, and author reps will be forming a company to carry out the objectives of this registry. I hope their vision isn't too conservative because delivery of a copyright database is too simplistic a solution given how our content businesses are developing. Visioning a comprehensive 'bibliographic' solution that marries uniform content identification with an end transaction is what our industry really needs. We don't really need another stand alone bibliographic database.

1 comment:

Eric Hellman said...

A few comments.
1. It's not a "google registry". The registry will be for the benefit of the plaintiff classes, i.e. authors and publishers.
2. Many of the benefits that could accrue from future adoption of ISTC are delivered today by things like OCLC's xISBN service and "FRBRization of book catalogs.
3. Since when are ISSN and DOI "not widely adopted"??? ISBN, ISSN and DOI all have succeeded because they have been clear in purpose- ISBN has been well managed as an identifier for a package relevant to the supply chain; ISSN identifies a serial name, and DOI identifies an article for hyperlinking.