Marek Niezgodka: Position Statement

From dml_wiki
Jump to: navigation, search

In order to create a future-proof WDML system, we should not be building yet another monolithic aggregator/digital library service. Instead I propose to focus on creating an open infrastructure, that would satisfy several conditions:

  1. Research publications are being increasingly analysed by machines for their semantics, relations between them etc. Given the increase of information produced and the advances in machine learning technologies, it can be assumed that this trend will continue and soon machines will be the primary consumers of research publications. This is of special importance in mathematics which provides much more space for formal analysis of the published contents. For this, WDML should be not only human-friendly but machine-friendly. It should also encompass linked data provisions, wherever applicable.
  2. The infrastructure should be open not only in the sense of open-source but also in its extensibility. To this end we should produce a common, component-based environment, using well defined interfaces for communication between individual modules. The modules, performing their specific functions, should be reusable but also replaceable with their alternative implementations. It should be possible to easily orchestrate the modules into individual applications, within an integrated but de-centralized, loosely-coupled layered architecture. If possible, the infrastructure should not be locked to a single technology.
  3. It should be carefully architected by a task force of computer scientists with experience in designing, developing and implementing large scale distributed digital library systems. Yes, not by mathematicians, or librarians but by software engineers.
  4. It should encompass much more content than the traditional peer reviewed literature. For first, open access repositories, and grey literature should be also included. Also provide hooks for including content from commercial publishers. And remain open for future types of content, particularly those, that can be used by machines. This in the future may include software, some data etc.
  5. It should take into account the emerging transition towards Open Access model of scientific communication, and particularly provisions for online collaboration in an open manner (e.g. wikipedia-like)
  6. It should provide support for long term preservation.
  7. Besides the technical solutions, the WDML should strive to build and maintain an active community around the infrastructure.
Personal tools
Namespaces
Variants
Actions
Navigation
Toolbox