Digital archaeology is increasingly shaped by how heterogeneous field records are modeled, linked, and preserved after surveys and excavations, in a constant effort to improve data availability and research capabilities. Here we are concerned with an innovative approach for treating specifically archaeological survey data, with future potential for the integration of excavation data as well. It should be noted that in general, archaeology is one of the disciplines in the humanities that inherently produces vast amount of digital data, and that as such it has been striving for digital solutions for their research-compatible publication for the last several decades – before Digital Humanities (DH) flourished, but belatedly in comparison to the natural and exact sciences, given that it shares with them many research tools and epistemological approaches (Trigger 2006).

By "survey data" we mean the diverse documentation generated during systematic site recording: coordinates and polygons, typological observations of archaeological remains and function, geological and geographic settings, artifact descriptions, photographs, orthophotos, videos, 3D models, field notes, dating attributions, expedition metadata, and bibliography (Banning 2002). Our approach is based on Linked Open Data (LOD), which, for archaeology, show how such materials can move beyond isolated spreadsheets or project databases into interoperable knowledge graphs (Schmidt et al. 2022). Specifically, we utilize the Wikibase platform for this task. We demonstrate its applicability, advantages and limitations through the case study of the Timna Valley Database Wikibase Project ( Ben-Yosef et al. 2024), a collaborative effort to model archaeological survey data in a structured, machine-readable, and open-access environment.

The Timna Valley, one of the world's oldest copper mining regions, serves as a field laboratory for studying the development of ancient mining and smelting technologies and human activity in a hyper-arid environment. Since 2012, the Central Timna Valley Project (CTV), led by Prof. Erez Ben-Yosef of Tel Aviv University, has conducted systematic surveys and excavations across the valley. Over the past decade, the CTV team has accumulated thousands of heterogeneous digital records - photographs, drone orthophotos, stratigraphic diagrams, radiocarbon results, GIS layers, and textual documentation - stored across disconnected folders and spreadsheets. The need for a sustainable, transparent, and interoperable platform therefore became urgent.

Our central hypothesis is that modeling an archaeological site as a semantic object rather than a static record changes both documentation and interpretation. Each site, artifact type, period, research expedition, bibliographic source, and dating method is represented as a Wikibase item connected by explicitly defined properties. The property set was designed against the requirements of more than 330 archaeological sites in Timna Valley (see Figure 1), with reusable authority items.

Our experience revealed that data modeling is both a technical and an epistemological act. Formalizing archaeological relationships made implicit assumptions visible: what counts as a site, how a period assignment is justified, when an expedition becomes an event entity, and how conflicting observations should coexist. This resonates with DH arguments that modeling is a knowledge-producing practice rather than a neutral technical wrapper (Ciula and Eide 2017).

The Timna Wikibase case shows how graph-based representation can extend DH beyond text-centric encoding toward infrastructures for material culture and landscape data. The Wikibase approach supports not only storage and retrieval but also comparison, visualization, public interpretation, and future extension to excavation data, laboratory results, and predictive analyses. It provides a potent platform for integrating past data as well as those that will be obtained by future expeditions, to produce a cumulative, always improving, interoperable open database, linked to global world-wide repositories of knowledge while maintaining local, case-specific integrity. It is also legible to interdisciplinary reuse – in our case, beyond archaeologists, it is relevant to environmental scientists, historians, heritage managers, developers, GIS specialists working on landscape features, modern mining in the region, and other aspects of human-environment interaction.

Beyond Timna, the methodology offers a replicable framework for archaeological survey projects seeking open, standards-aware, and sustainable data infrastructures. Importantly, modeling survey data is not merely a technical afterlife of fieldwork; it is a scholarly act that turns documentation into living, connected knowledge, one that requires collaborative efforts of the broader community.

Figure 1. Map of Timna archaeological sites generated from a SPARQL query over site coordinate statements.