ANAMED’s Linking Anatolian Pasts (LAP) project aims to establish and sustain an interactive online portal that brings together now “siloed” research collections and datasets pertaining to the Anatolian past including aerial imagery, photographs, maps, manuscripts, archives, publications, museum objects, whole archaeological databases, and more. Rather than attempting to consolidate an increasingly large number of relevant datasets at one institution or on one centralized server—essentially creating yet another silo of research datasets—the project takes a decentralized approach that virtually centralizes datasets through a Linked Open Data (LOD) framework, following a “set of best practices for publishing and interlinking structured data for access by both humans and machine” (World Wide Web Consortium, 2014, Scope). Through adoption of an LOD model, ANAMED aims to build a network of partners both within and outside Koç University who are willing to expose currently existing digital datasets in their collections to targeted linking and spatialization, eventually enabling textual, image, and spatial querying across diverse datasets unified by a common geographic framework: Anatolia.
As such, the LAP project builds on the success of and resembles the earlier Koç University-led community-linking project of biblioPera, the “Beyoğlu Research Centers Network,” which serves as an information sharing platform for partner research centers, with the primary goal of making dispersed library catalogues accessible via linking through a single online portal. Rather than focusing on library collections, however, the LAP project supports social-science and humanities research into Anatolian pasts by serving as an interactive research tool, where primary data and metadata (e.g., material collections, places, images, and objects) can be queried via geography, keywords, specific metadata, and/or image matching.
Just as the LAP project aims to serve as a gathering point and suite of research tools for otherwise unconnected and dispersed datasets, so too will it draw from and contribute to other LOD portals featuring geographically broader sets of data. As such, it may interact most directly with the well-established Europeana, the Mediterranean focused Pelagios Commons and its “Linked Pasts Network” working group, as well as established vocabularies for things like place names (Geonames and Pleiades) and chronological periods (PeriodO). Other major institutions adopting LOD frameworks include the British Library, National Library of Spain, and US Library of Congress, as well as the Online Computer Library Center and the Getty Research Institute.
Required partners within Koç University include (at least) the IT Directorate (KU IT), the Suna Kıraç Library (SKL, and ANAMED Library); and the Artificial Intelligence Research Center (AIRC).
The project revolves around four interrelated work packages:
Package 1: the Archival Anatolian Atlas (or Historical Aerial Imagery Atlas?)
The Archival Anatolian Atlas (AAA) project aims to establish the geospatial foundation on which LOD datasets can be displayed and queried: an interactive web-GIS with layer-toggling as well as tools for textual and spatial queries as well as point, line, and polygon digitization for download. The foundation will consist of a Turkey-wide set of georeferenced 1950s aerial imagery, revealing Anatolian landscapes as they were primarily before the onset of industrial-scale agricultural, rural development projects, and the explosive growth of urban areas. The atlas will include also limited focus areas of earlier aerial imagery as well as historical digital elevation models, and a variety of more recent and contemporary satellite and aerial imagery, maps, and digital elevation modes. As partners join the LAP project from within and outside the Koç University community, the aim is to display and link to the original repositories of a variety of dispersed datasets.
Package 2: Linking and Spatializing Digital Collections
A): LAP Partners within Koç University
In partnership with the Suna Kıraç Library, an initial goal will be to link selected parts of KU SKL’s Digital Collections via protocols to make CONTENTdm data LOD-ready and to spatialize them for geographic display in the AAA. Photographic collections that may be particularly appropriate for this project include but are not limited to those of Josephine Powell and Hatice Gonnet-Bağana. The Istanbul Soundscapes, Ulla Johnson, and Nigizberk collections may be similarly appropriate, as may be the collections associated with ANAMED’s peer research centers AKMED, GABAM, and VEKAM. The georeferenced map collection of M. Erdem Kabadayı’s UrbanOccupations project, too, may eventually make it a suitable partner in the LAP project, as might similar projects conducted under the Koç University umbrella.
B): LAP Partners outside Koç University
Simultaneous with the linking and spatializing of collections and data already digitized and housed within Koç University-affiliated collections, the LAP project aims to expand its partner network to any and all institutions willing to expose their collections to the LOD network. Relevant collections may focus primarily on Istanbul (e.g., SALT’s Pervititch Maps, ARIT-SALT’s American Mission Board archives, and more, the İstanbul Araştırmaları Enstitüsü, the Pera Müzesi), while others may have broader foci (e.g., the Levantine Heritage Foundation, the Levantine Ceramics Project, the Eski Türkiye Fotoğrafları Arşivi, NIT-Kiel archive, Kültür Envanteri Atlası, the Hrant Dink Foundation’s Turkey Cultural Heritage Map). This project could attract many other possible partners and collections. For other Ottoman and Turkish period possibilities, see the recent overview of digital sources compiled by Nicole van Os (Leiden University Turkish Studies).
Package 3: Linking, Spatializing, and Making AI-Searchable Digitized Archaeological Collections
Through a proposed partnership with the Müzeler Dairesi of the Cultural Heritage and Museums General Directorate of Turkey’s Ministry of Culture and Tourism and Koç University’s recently established AI Research Center, the LAP project aims to develop an online tool that will significantly improve the ease, speed, and accurate reliability of object and image identification among museum staff, archaeologists, art historians, and other researchers and cultural heritage managers. By exposing the digitized inventory of “envanterlik” museum objects, their associated characteristics, and metadata via LOD protocols (Müze Dairesi) and by collating previously published and digitized images, characteristics, and metadata for the same into structured and openly linked datasets (excavation and museum catalogues), an enormous and machine-readable reference database will be compiled. Taking advantage of the latest developments in image matching algorithms using AI concepts (KU AI Research Center), the proposed tool will enable keyword and/or image searches that return the best matches for entered parameters. While AI algorithms will take advantage of the entire database to generate results, only data (textual or image) for which permissions have been granted will be viewable as part of returned results. That is, the system enables AI image matching, keyword querying, and display of distributional data without having to actually expose the comparative reference data itself to the public. Similarly, provenience resolution can be set by the contributing partner, depending on cultural heritage preservation concerns, for instance, to the province, museum, site, or specific context level.
Package 4: LAP and Living Archives
The final work package of ANAMED’s LAP project is to support (if not develop anew) hosting capabilities for dynamic and queryable archaeological research databases. The envisioned pilot database for which negotiations are underway is the Living Archive of Çatalhöyük (LAÇ). The LAÇ is an online research portal for the 25-year Çatalhöyük Research Project (ÇRP) directed by Ian Hodder that aims to be a dynamic repository that can be used to display, query, and download existing aspatial and spatial datasets and that can receive new or modified datasets, hence the “living” in the title. Similar online data networks for Anatolian archaeological projects have been developed also by scholars at the University of Chicago (OCHRE, D. Scholen, S. Schloen, et al.) as well as the University of Toronto (CRANE, T. Harrison et al.), and others are under current development (e.g., for the Central Lydia Archaeological Survey (CLAS) and Kaymakçı Archaeological Project (KAP), C. H. Roosevelt, C. Luke, et al.). By engaging with these and perhaps other Koç-supported projects (e.g., Sagalassos and its Sagalassos Integrated Information System (SIIS)), and by demonstrating the potential of openly publishing archaeological research databases and results with LOD, in machine-readable ways, the project expects to attract requests for hosting, if not guidance on how other datasets can be made available in similar ways.