Scott Markel is Vice President, International Society for Computational
Biology (ISCB); Co-Chair, ISCB Publications Committee; and Associate Editor,
The BioLINK SIG meeting has been regularly held in association with the ISMB
conference (Intelligent Systems for Molecular Biology—the annual
conference of the International Society for Computational Biology) since 2001,
focusing on the development and application of resources and tools for biomedical
text mining. The SIG (Special Interest Group) is interdisciplinary in nature, and
brings together researchers applying natural language processing, text mining, and
information extraction and retrieval in the biomedical domain with scientists from
bioinformatics and biology. This year's meeting at the combined ISMB/ECCB
(European Conference on Computational Biology) conference in Stockholm includes two
new sessions, one dedicated to extraction of information from images, and one
devoted to the future of scientific publishing. The publishing session, co-organized
by BioLINK with the collaboration of the ISCB Publications Committee (
Last summer, we undertook manual semantic enhancements to a biomedical research
article, providing enrichment to its content and increased access to datasets
within it, to provide a compelling existence proof of the possibilities of
semantic publication (
In my presentation, I will explain what we achieved by means of a live link to the online enhanced paper, discuss the significance of this work in terms of recent developments in automated text mining, and consider the future of semantic publishing as part of mainstream research journal production workflows. My aim is to excite the imaginations of researchers and publishers, stimulating them to explore the possibilities of semantic publishing for their own research articles, and thereby break down present barriers to the discovery and reuse of information within traditional modes of scholarly communication.
I will discuss a number of initiatives in which I am involved to improve and
enhance access to scientific knowledge from collections of research articles.
First, at Elsevier Labs, we added manually annotated Structured Digital
Abstracts in
Scientific literature is nowadays distributed in electronic form through online
Web portals. ELIXIR Work Package 8 (WP8;
Scientific literature is kept in national and international repositories that currently still lack connectivity. The biomedical community is driven by the idea of the integration of all data resources (including literature) from the level of molecular biology to medicine, leading to multidisciplinary research. The appropriate infrastructure and tools need to be in place to facilitate full exploitation of the literature across scientific domains and at various levels of end user expertise. Scientific literature is unstructured in contrast to the scientific databases. This has led to (1) the development of text mining and knowledge discovery solutions that recover facts from the scientific literature, (2) curation efforts to include scientific facts into the main databases, and (3) efforts around various wiki-like projects to produce annotations. The exploitation of the scientific literature has to (1) fulfill multidisciplinary needs, (2) exploit ontological resources (Semantic Web approaches), (3) deliver enhanced digital content, and (4) follow standards for efficient integration.
Scientists (at least their profiles) and their scholarly output exist in cyberspace, but the relationship between the two is far from established. Scientists may not be identified uniquely, and much of their output is not easily referenced. The Digital Object Identifier (DOI) was a big step in uniquely identifying a scientific journal publication, and has been embraced by the majority of publishers. I would argue that the time is here for extending this scheme to uniquely identify scientists (authors) with all their respective scholarly output. This is much more than traditional journal publications, and includes database depositions, reviews for grants and journals, blog postings: in fact, anything they wish to have uniquely associated with their name. I will discuss efforts in this direction and what I think it will take to really make such a scheme work—a scheme that starts with the publishers.
The publishers' panel will follow the scientific presentations. The
publishers will be free to comment on the presentations or to address other
topics, such as validation processes and quality measures (e.g., the future of
the peer review model, alternatives to impact factors), dissemination (e.g.,
open-access models), and discoverability (e.g., linking, applying new
technologies). Confirmed participants include Claire Bird (Oxford University
Press), Mark Patterson (Public Library of Science), Matt Day
(
The BioLINK SIG meeting will be held at the Stockholm ISMB/ECCB 2009 meeting on
Sunday and Monday, the 28th and 29th of June. The Future
of Scientific Publishing session will take place in the afternoon of Monday, the
29th of June. See