A review of Recogito, an open source platform for semantic annotation, maintained by Pelagios Commons
Christy Hyman, University of Nebraska - Lincoln
Recogito is a free, open source Web platform for semantic annotation—the marking up of text, image, or tabular documents according to controlled vocabularies. Through an easy-to-use interface, users can identify places in texts, scanned maps, or databases and disambiguate them by linking them to global authority records (gazetteers); identify other entities (like people) and use tags to create a flexible systematic typology; export the annotations into other applications using a range of data formats; and display the place annotations on inbuilt maps. Users can work alone in a closed workspace, together as groups of collaborators, or in a public setting using Recogito as a crowdsourcing platform. Recogito keeps track of version history and edit provenance, and it can be customized with different name authorities for geo-resolution. It is also easy to apply Named Entity Recognition (NER) to texts, with the option of choosing between different recognition engines and gazetteers.
Recogito has been developed under the umbrella of the Pelagios project, in three phases funded by the Andrew W. Mellon Foundation. To support Pelagios’s aim of establishing a simple and effective method of linking online resources, Recogito was initially piloted for internal use by the project team to show how research on the ancient Mediterranean world and late antiquity could be transformed by semantically annotating place references, if enough resources could be connected. Such was the interest among scholars and teachers from a range of other disciplines—including Chinese, Ethiopian, and Islamic traditions, as well as early modern Europe and the Americas—that the Pelagios team redesigned the tool for general use by the subject specialist without need for coding skills. This has meant focusing on usability; providing support for general annotation and commentary (not just geo-resolution); and enabling users to create their own accounts and manage their own personal collections and collaborator groups. The development and architecture of Recogito has been driven by the principles of decentralisation and open source software; the tool creates semantic links between collections, rather than imposing an overarching architecture. All software outputs and documentation are openly licensed and released via the Pelagios GitHub Repository.
Recogito is now being maintained as a partner of the new Pelagios Network, an open, non-profit association advocating for and enabling the production and use of linked open geo-data in the humanities and cultural heritage. Current strategy has identified two key areas of growth: community and technical development. Regarding community development, the Pelagios team runs Recogito workshops for galleries, libraries, museums, and archives, with a view to better enabling curators to enrich their collections. Technical development focuses on integrating options with external sources and developing standards, such as the International Image Interoperability Framework (IIIF) or Text Encoding Initiative (TEI) markup published through the Clear to Send (CTS) protocol. Among its successes, Recogito has over 4,200 users, who have made almost 3.5 million annotations; members of the Digital Humanities community voted it Best Tool of 2018 in the Digital Humanities Awards; the Open Knowledge Foundation funded workshops to use Recogito in undergraduate modules; and EU-DARIAH is integrating Recogito into their own suite of scholarly digital tools.
Scholars who are engaged in spatial analysis often find themselves in search of maps and other documents containing geographically referenced data that will aid their investigations of human connections to past landscapes. When those sources are acquired, however, there is still much work to be done to extract spatial data for further analysis. Recogito, a free and open source annotation tool, responds to this need by making it easy for researchers to identify, record, and export spatial information from an array of sources. Semantic annotation, Recogito’s most innovative feature, sets it apart from other tools, such as Delicious and Flickr, that focus on storage, annotation, or tagging. Recogito allows researchers the opportunity to not only identify places in texts, scanned maps, or databases, but also to disambiguate spatial data by linking it to global authority records such as gazetteers.
Semantic annotation involves a process where user-supplied information — an array of concepts such as people, things, and places — is attached to source content. Recogito provides users with a dynamic workspace for transcribing and annotating sources that are rich in spatial data while serving as an online center for projects requiring collaborative input. This tool guarantees an ease of use that promotes further information discovery of geographic sources. Recogito has been developed under the auspices of the Pelagios, a formal network of partners whose work promotes linked open data between common references and places. Recogito was funded by the Andrew Mellon Foundation, and since its initial release in 2015 has over 5,013 users who have made over 4.43 million annotations.
The registration process is quick and simple. Researchers may upload texts and images, create notes and tags, and export data to other tools. New users will find the tutorial helpful for site navigation. Users will be impressed with the site’s attractive layout. The color scheme emphasizes readability with its light array of hues contrasted with black and grey text in a modern font style. There are options for making work sharable, allowing for reuse and collaboration. Researchers desiring a central platform that allows collaborators at multiple institutions to work on an image will find that Recogito meets those needs.
Recogito has been used by scholars worldwide in a range of historical studies. A challenge that may arise with the tool involves the linguistic representation of certain toponyms. In the case of a project based at Boğaziçi University in Istanbul, Turkey, Ottoman place references were not widely represented. Antonis Hadjikyriaco, the scholar heading the project, applied for and received funding for work extracting toponyms from Ottoman historical sources. Hadjikyriaco remarked that in working with Recogito scholars find valuable lessons in spatial history methods while exploring some of the difficulties of verifying the reliability of historical data.
Human annotation practices complicate what might be expected from a tool as the political implications of place naming are independent of existing data authorities. In this way, Recogito is a fine example of the possibilities for future tools based on open source development, where scholarly inquiry leads to greater access to expanded paths of knowledge.
 Simon, Rainer, Elton Barker, Leif Isaksen, and Pau De Soto CaÑamares. "Linked Data Annotation Without the Pointy Brackets: Introducing Recogito 2." Journal of Map & Geography Libraries 13, no. 1 (2017): 111-32. doi:10.1080/15420353.2017.1307303.
 Helen J. Davies and Alexander Zawacki, “Collaboration and Annotation: Pelagios, Recogito and Multispectral Imaging of Cultural Heritage Objects,” Europeana Pro, no. 12 (March 2019), https://pro.europeana.eu/page/issue-12-pelagios#collaboration-and-annotation-pelagios-recogito-and-multispectral-imaging-of-cultural-heritage-objects.
 Antonis Hadjikyriaco, “Using Recogito in the (Ottoman History) Classroom,” Europeana Pro 12, no. 12 (March 2019), https://pro.europeana.eu/page/issue-12-pelagios#using-recogito-in-the-ottoman-history-classroom.
 Pelagios Network provided the funding for the Hadjikyriaco’s project.
 Hajikyriaco, “Using Recogito in the (Ottoman History) Classroom.”