This grant funds the continued development of SHARE, an open access database and service that links together university-based data repositories in an effort to make scholarly research widely accessible, discoverable, and reusable. Developed in collaboration between the Association of Research Libraries and the Association of Public and Land Grant Universities, SHARE includes not only a searchable database of research, but also a scholarly research notification service that allows users to keep abreast of new developments in scholarship, for example, when a relevant new white paper is uploaded by a scholar they are following, an important dataset is updated, or a previously unpublished study is published. Funds from this grant will support the continued development and expansion of the SHARE platform, including efforts to increase the number of participating data providers, integration of SHARE into the diverse workflows of member institutions, and the cleaning and normalizing of the, oft-messy, metadata that powers the SHARE search algorithms.

Community engagement managers are increasingly seen as vital and irreplaceable elements for the smooth functioning of healthy online communities. Though it is a new field, community engagement has matured quickly, with a growing body of common methods and best practices. Individuals playing this role in scientific contexts, however, are often isolated from this community of practice, and left to trial and error to figure out how to be most effective. This grant supports a Community Engagement Fellowship program at the American Association for the Advancement of Science (AAAS) for a network of community engagement managers that will connect several scientific fields. The Fellows will be based in a combination of AAAS-affiliated scholarly societies and large multidisciplinary collaborations, which will build capacity in scientific organizations. In addition, fellows will be brought together for annual training boot camps and monthly professional development webinars, allowing them to share ideas, common challenges, and best practices. Grant funds support approximately half of the planned 18 fellows of the initial cohort, with additional funds provided to offset the costs of outreach, fellow selection, and program administration.

Mathematical formulas are undiscoverable by modern search engines. If you are looking for a famous theorem or an equation with a name, standard search engines like Google or online encyclopedias like Wikipedia can direct you to it. But if what you are looking for is an equation that expresses one variable in terms of another, you are out of luck. Because the consumer base for such information is small and because the task of programming computers to recognize mathematical formulas is difficult, no major search engine has prioritized mathematical search. Yet from a societal point of view, the benefits of accelerating discoveries by providing such search capabilities could surely be enormous.   This grant funds a project by John Lafferty from the University of Chicago and David Blei from Columbia University to advance the field of mathematical search by developing a software program that uses sophisticated pattern recognition and statistical machine learning techniques to recognize and identify mathematical formulas on the web.

The sustainability of scientific data repositories is a matter of much concern. One recent success story is the Arabidopsis Information Resource (TAIR), a plant biology database that shifted from a grant-funded to a community-funded business model through the strategic development of a discriminating paywall that grants free, limited access to many users while requiring a sliding scale paid subscription for full, unlimited access. TAIR Director Eva Huala believes that a more flexible version of this paywall software could enable many other data repositories to develop their own variations on this model. While there are several for-profit startups offering such services, none offer the functionality needed by scientific data repositories, and these repositories appear to be much too small a market to draw those startups' focus. Funds from this grant will support Phoenix Bioinformatics, the 501(c)3 that runs TAIR, in its efforts to develop a flexible, portable version of its paywall software that could be used by a wide variety of scientific data repositories.