ImpactStory (previously Total-Impact) allows researchers and organisations to gather a wide range of impact metrics about multiple forms of scholarly output. Users create collections of research objects – articles, data sets, websites, presentations, etc. For each object, the system searches across a number of data sources to collect indications of impact and aggregates the information into a distributable report.
Development began at the 2011 Beyond Impact Workshop and continued with funding from the Open Society Foundation via the Beyond Impact project. Current development is led by Cottage Labs, with help from individuals at the University of North Carolina and Duke University and funding from the Alfred P Sloan Foundation.
Licensing and cost
Open-source MIT license – free.
The service is currently under active development. The source code is available from GitHub.
Platform and interoperability
The ImpactStory service is web-based and therefore platform agnostic. For local installations, the project recommends either a clean Ubuntu 10.04 server or a virtualised server using Virtual Box. The code is mainly Python 2.7, with some scripts written in Perl. Metrics draw from a number of data sources: CrossRef; Mendeley; Slideshare; Dryad; PLoSALM; PLoSsearch; Facebook; CiteULike; Wikipedia; Delicious; PubMed; Topsy; Research Blogging; GitHub; and Sourceforge. The project offers an API for those wishing to incorporate further information sources.
The service collects information about multiple data types: published articles via their DOIs, PubMed IDs, or Mendeley UUIDs; data sets through their Dryad DOIs or accession numbers from GenBank, PDB, Gene Expression Omnibus, and ArrayExpress; software through its GitHub or SourceForge URL; slides through their SlideShare URL; and any other output that has a URL. ImpactStory can currently aggregate over 50 metrics such as the number of Mendeley readers of an article, a presentation’s number of Slideshare downloads, and the number of Facebook users who ‘liked’ a post about the object. The project acknowledges a number of limitations, arising from the difficulty of deduplicating objects with multiple identifiers, and locating objects via their identifiers. There is a limit to the number of items that can be included in a single report.
Documentation and user support
The website offers a FAQ and links to an explanation of the principles underlying and motivating the service. The project supplies documentation for creating plugins to integrate additional sources of information. Further user support can be found through an active Google Group, a blog, and a Twitter a ccount.
ImpactStory uses an extremely simple web interface. Interpreting reports can be challenging; there are a high number of metrics displayed, and as the field of alt-metrics is still in its infancy, a lack of context and comparison leaves the true importance of many of the statistics opaque.
Experience with social media is helpful for interpreting the reports.
ImpactStory relies on resource identification standards for compiling their reports. The DOI system is currently the most widely used, but a lack of standards is one of the core problems of the endeavour.
Influence and take-up
The project Twitter account has over 1500 followers; further information is unavailable.