TDWG 2013: SCAN – Leveraging Filtered Push Technology to Enhance Remote Taxonomic Identifications
Presentation slides are now posted for the TDWG 2013 SCAN talk.
Date: 2013-10-29 02:45 PM – 02:50 PM
SCAN – the Southwest Collections of Arthropods Network (http://symbiota1.acis.ufl.edu/scan/portal/) – is the first regional arthropod biodiversity data network that utilizes the Symbiota software platform (http://symbiota.org/tiki/tiki-index.php). Since its origin in 2012 SCAN has unified and newly created specimen-level occurrence records on-line pertaining to nearly 15 south-western United States arthropod collections; including more than 515,000 records that represent some 18,000 species. However, due to the disproportionately mismatched diversity versus taxonomic expertise for the region and focal taxa, at least one third of the specimens are not identified (authoritatively or otherwise) to the level of species, with concomitant limitations for derivative taxonomic or evolutionary/ecological research. The member collections are typically separated from each other geographically by distances that prohibit frequent interactions with regional or global experts, except in virtual realm. SCAN has therefore implemented a Filtered Push (FP) based service (http://wiki.filteredpush.org/wiki/) whose primary purpose is to connect high-quality imaged of yet insufficiently identified specimens with suitable experts who can provide identifications remotely. This is achieved through the FP-server system which both records these contributions externally and pushes them back into the source Symbiota platform for review, acceptance, or rejection by the respective collection/node leaders. SCAN is therefore primed to utilize FP at a large scale and with a well circumscribed focal purpose that is relevant to the specific needs of this collections network. We illustrate the SCN/FP workflow, underlying concepts and technology, and current state of implementation and usage. FP allows experts to gradually accumulate credit and “reputations” for their identification contributions, and thus represents a promising means to improve data quality through transparent and distributed expert involvement and attribution.