Archeologia e Calcolatori 34.2 - 2023

The Authors propose a knowledge map to analyse and access scientific contents related to Digital Archeology by leveraging various Machine Learning (ML) techniques. The case study concerns the articles published in our international journal «Archeologia e Calcolatori» in the decade from 2011 to 2020 and, as a benchmark, the publications in the ‘Computer Applications and Quantitative Methods in Archaeology’ (CAA) conference proceedings and journal. The titles and abstracts of the publications featured in these two data sets were analysed using a supervised classification approach into the subfields of computer science, based on the ACM’s taxonomy, and by applying topic modelling techniques to discover emergent topics, Named Entity Recognition to identify specific archaeologically relevant entities, and geotagging techniques to link articles with the geographical locations they discuss. The results achieved, although preliminary, provide some methodological suggestions: i) the opportunity to build custom analyses by taking advantage of the increasing availability of open data and metadata; ii) the scope of the contribution of archaeology, and in particular of computational archaeology, to the Heritage Science interdisciplinary domain; the heuristic and predictive role of different ML techniques to gain a multi-faceted access to data analysis and interpretation.

Large dwelling spaces, characterised by a continuous human occupation and for different practices, represent crucial archaeological contexts in reconstructing the organization of production and consumption activities within prehistoric communities. However, the archaeological record related to such depositional contexts often appears spatially disordered and dominated by a chaotic distribution, the result of the interaction of human and natural agencies over time. On this matter, computer modelling offers a wide range of methods to disentangle the apparent spatial chaos and assess the dynamics behind the distribution of the remains, both those deriving from human activities carried out on the spot and those resulting from later disturbances. In this framework, one of the main issues is the reconstruction of the complex of materials and tools from some human activity. This paper explores the effectiveness of Gcross function analysis to investigate dynamics of interactions between different categories of remains in a large dwelling space, addressing the question of how each category of remains interacts in the space with the others. As a case study, the analysis focuses on a wide area within the Bronze Age fortified settlement of Coppa Nevigata (Southern Italy).

This research aims to use quantitative and repeatable GIS techniques, as well as Machine Learning algorithms, to study the settlement patterns in Southern Etruria during the final phase of the Bronze Age (1150-950/925 BC). The region of Southern Etruria is located in present-day Latium, Tuscany, and Umbria. The study, which includes 166 settlements, focuses on identifying the morphological characteristics of these settlements by means of raster analysis. Using a Machine Learning approach, the research will compare real settlements with random points within the region to understand the specific characteristics of the settlement pattern in the landscape. The study will also examine the use of feature selection and features importance methods to select the most significant features of a multivariate dataset.

This article aims to quantify the rate of accidental ceramic discard on the archaeological site of Siponto (Italy), where in 2022 the University of Bari and the University of Foggia conducted fieldwork and training for students at different education levels (BA to PhD). The goal was to identify and quantify factors leading to the accidental discard of ceramic sherds by excavators on the spoil heap. As a pilot project, a few variables have been considered to count the minimum number of individuals found after sieving soil composition and colour, weather conditions, time variables, sherds size, colour, and vessel part. Other categorical or presence/absence variables have also been considered. This enlightening investigation shows the bias in post-excavation quantification of ceramic finds. Results indicate that 30% of the fragments of pottery retrieved from the spoil heaps, used in this experiment, were diagnostic. The study also helps the educators on-site to identify the types of vessels that might be less clear for the students.

Mountain archaeology has a long research tradition and in recent years the number of studies on this topic has increased considerably, shedding new light on the dynamics of mountain’s communities. Versilia and Garfagnana districts (Lucca, North-Western Tuscany) largely fall between the Apuan Alps and the Apennine ridge. Although these territories have never been systematically investigated, the collection of all available archaeological legacy data indicates a settlement pattern of undoubted interest for the Roman times. This paper aims at exploring the settlement pattern of these mountain territories, integrating Point Pattern Analysis and Logistic Regression to achieve a predictive map of archaeological presences and to analyse their interrelations with the environment. Analyses prove the spatial dependence of finds with geomorphological and pedological variables, but also with the distance to major watercourses and solar irradiation. Based on the considered variables, the predictive map confirms that the foothill and gentler slopes facing E-SE areas have suitable characteristics for permanent settlement. Moving towards the more inland and higher altitude territories, the non-event areas increase proportionally, especially along the ridges, and the steeper, north-facing areas. Thus, the results make it possible to integrate the archaeological framework, clarifying human-environment dynamics, and directing new studies.

The prehistoric village of Portella (Salina, Eolian Islands, ME), occupied during the Middle Bronze period, was discovered by chance in the 1950s and excavated in three different phases, from the moment of discovery until 2008. The site is well preserved because, after partial destruction, it was rapidly covered by landslides of volcanic rock. That notwithstanding, the erosive activity had negative effects, too: the lateral erosion of the ridge is destroying part of the village; possibly a portion was already collapsed at the time of discovery. We can consider this erosion an unstoppable process that needs programmatic, continuous monitoring of the site. In the article, we will discuss how it is possible to make one ‘screenshot’ of the monument’s status with the help of new technologies, specifically, photogrammetry through UAVs. To enhance fruition, a virtual tour of the site was also created. This allows people with physical disabilities or reduced mobility to access, though virtually, the site. In conclusion, other possible areas of application of this low-cost and expeditious methodology are suggested, in particular inaccessible or overcrowded sites.

The article systematically explores two processes of virtual reconstruction of archaeological contexts: ArchaeoBIM and Extended Matrix. The focus is on the theoretical frameworks behind their development, the proposed operational processes, and the products derived from both methodologies. The informative potential of the virtual models resulting from these reconstruction processes will be discussed, as well as the application-related issues. A substantial part of the article will be dedicated to the development of an integrative protocol aimed at incorporating the informational structure of the Extended Matrix within an ArchaeoBIM model. The process has been applied to the case study of House 1 in Regio IV, Insula 2 of the Etruscan city of Marzabotto, which was excavated in recent years (1988-1998) and thoroughly documented. Final considerations are then directed towards future development prospects and the integration of this virtual product within a Geographic Information System.

The Messapian necropolis of Monte D’Elia (used from the 6th to the 2nd cent. BC) is about 300 m S of the modern town of Alezio (Lecce, Apulia), which lays on one of the most important ancient settlements of the Salento peninsula (ancient Messapia). The site was investigated between 1981 and 1985 by the Apulian Archaeological Superintendence. In 2020, archaeological investigations have been resumed by the Department of Cultural Heritage of the University of Salento, with the participation of researchers from the Institute of Heritage Science of the National Research Council. Their research activities focused on geophysical prospecting and topographical surveys, performed thanks to the combined use of laser scanner and digital photogrammetry by drone. This contribution illustrates these research activities, which are aimed at understanding the general plan of the necropolis, through the integration of the information published in the 1980s with data from new investigations and surveys, and, more generally, at the reconstruction of its topographic organization and extension.

The ISPC-CNR is conducting research activities aimed at identifying ancient quarries that supplied the public building sites in the Athenaion of Castro, where excavations revealed dry-stone city walls made of calcarenite blocks and built in three phases between the 4th and 2nd cent. BC. The paper is focused on the large coastal quarries of Porto Miggiano, located approximately 4.5 km NE of Castro, where the extraction was carried out in function of sea transportation of the blocks. The extraction sites were documented through laser scanning and photogrammetry by a drone. In the main quarry, which covers an area of approximately 0.4 hectares, thanks to the metrological analysis of the stepped faces, four extraction areas were identified corresponding to at least five cultivation phases. The oldest and most extensive one affects the central sector of the quarry, where large blocks compatible (for dimension and macroscopic aspect) with those of the third phase of the Castro city walls, dated to the beginning of the 2nd cent. BC, were extracted.

This paper presents the results of the research undertaken through a series of on-site surveys and studies (2019-2022) of the site of Castellito di Ramacca (Catania, Sicily). The site, located on the top of a low hill (106 msl), was partly excavated in the late 20th century and interpreted as a Roman rural building, possibly with a special function (road station). Its name is also attested in ancient sources (Capitoniana). The site was re-evaluated using various non-destructive prospecting methods (aerial photogrammetry, fieldwalking survey, architectural recording), precise location with RTK GNSS and integration of the legacy data in GIS. This approach confirmed a new addition to the already known villa complex and contextualized it as a focal point of the rural settlement system. Based on the topographical data obtained from the survey campaign, conducted by integrating of different techniques, we propose a reconstruction of the villa’s elevation at its peak in Late Antiquity. This study illustrates the great potential of applying digital technologies for a new understanding of Roman villa remains.

The paper presents the results of a two-year archaeological survey carried out in the Iraqi Kurdistan, namely within the Navkur Plain that has been extensively explored by the University of Udine since 2012. The surveys were planned in advance using Remote Sensing products available online and processed with Google Earth Engine, a large-scale cloud computing service specifically designed to process geospatial big data and especially satellite imagery. Images from Landsat 5, Landsat 7 and Sentinel-2 platforms were selected, processed and assessed. After two years, an overall number of 46 new and previously unknown sites have been localized and surveyed, contributing to the knowledge of the past history of this portion of the Kurdistan region and testing the use of Remote Sensing cloud-computing applications in the context of Near Eastern archaeological research.

The nucleation of the rural population was a widespread phenomenon during the Middle Ages that interested many areas of Western Europe. However, many of these sites are now deserted with the underlying phenomena causing these abandonments not always easy to reconstruct. Archaeologists have been interested in these medieval settlements since the middle of the 19th century, and remote sensing has played a decisive role in mapping hundreds of them. This also applies to many parts of the Po Valley but not the Romagna plain, where hundreds of medieval sites are known but almost exclusively based on written sources. However, the increasing availability of aerial and satellite images offers a valuable opportunity to bridge this knowledge gap. The systematic study of legacy images allowed the mapping of new defensive elements and reconstruction of the general plan of six deserted medieval fortified settlements in the broader hinterland of Ravenna. PlanetScope 3m resolution images were later exploited to continuously monitor these sites during periods prone to crop marks formation to detect the presence of wide crop/soil marks (e.g. ditches). Six successful field verifications demonstrate that these ‘coarse’ images are sufficient to plan drone surveys that can allow the mapping of additional smaller features.

Data acquisition digital methodologies have become a reliable tool for surveying buildings with heritage values. Laser scanning has become the preferred method for performing 3D digital surveys because of its high accurate results; even though, the cost associated with it is usually high. Emerging technologies have been able to produce low-cost data acquisition methods, and they are currently being incorporated as part of digital survey projects. Using the ‘Domus of the Calendar’ as a case study – an exceptionally unique archaeological and architectural site that was incorporated to the structural foundation system of the basilica of Santa Maria Maggiore – the present investigation aims to evaluate the data quality of two low-cost emerging technologies, namely SfM (Structure from Motion) and the iPad LiDAR system. This evaluation was developed by comparing low-cost technologies data acquisition capabilities with those of the laser scanner. The data for this test was obtained during an integrated survey campaign aimed at executing a critical analysis of the many historical layers of the Santa Maria Maggiore basilica. The results obtained from this investigation highlights the reliability of the different techniques implemented and suggest a useful solution for different and recurrent multi-scalar contexts.

The poorly preserved human bones discovered during the DHOMIAP Project excavation of the necropolis KR-N1 in the area of Khor Rori (Dhofar, Sultanate of Oman) were an opportunity to apply, for the first time in Dhofar’s pre-Islamic funerary contexts, 3D photogrammetry to osteo-archaeological studies. The low economic engagement and the execution speed make this technique essential in the documentation of barely accessible archaeological remains and contexts, as already witnessed by previous studies conducted outside this research area. This paper aims to find a more appropriate method and setting for 3D model photogrammetric reconstruction of human remains, demonstrating the importance of this digital technology for the study of poorly preserved osteo-archaeological remains. For these purposes, the results obtained using two different settings for image acquisition (one with macro and one with standard lens) were compared and discussed.

The MAGOH (Managing Archaeological data for a sustainable Governance of Heritage) project is a two-year project funded by Regione Toscana, co-funded by the Italian Ministry of Culture (MIC) and coordinated by MAPPA Lab of the University of Pisa. The project was designed to address the needs of the Superintendencies of Florence, Pistoia and Prato and of Pisa and Livorno to manage archaeological data. The project represents the development of the MAPPA project on a larger geographical area of 72,000 km², corresponding to almost all of Northern Tuscany. MAGOH system is composed of a web-based back-end which allows collecting textual and vector data and the archaeological documentation. It contains around 8000 archaeological interventions openly accessible through the web platform and reusable as open data following FAIR principles. Furthermore, through an appositively developed API, it is entirely interoperable with GNA, the National Geoportal for Archaeology, managed by the MIC.

In 2005-2006 the excavations in the Sanctuary of Apollo conducted by the Italian Archaeological Mission in Hierapolis in Phrygia (MAIER) brought to light about sixty fragments of a larger-than-life marble statue of Athena. This paper presents the discovery, highlighting the role played in the research by digital technologies, especially 3D modelling and reconstruction techniques, the application of which mitigated the highly fragmentary nature of the evidence. The first section of the paper highlights the importance of the context of discovery of the fragments, which were found, together with other sculptural and architectural elements, in front of Temple B, in a deposit of discarded material related to the destruction in situ of part of the sanctuary’s decorations in the Byzantine period. The second section describes the plan drawn up by the MAIER to publish a comprehensive scientific description of the sculptures of the Sanctuary of Apollo. The project aims to reconstruct the sculptural decoration of the sacred area in the Imperial period, combining information on the types of statuary and the iconographic subjects and themes with excavation data and the epigraphical documentation from the sacred area. The project includes the study of the statue of Athena, for which this paper provides the description and the results of the historic and artistic analyses. The final section is centered on the process that led from the creation of the digital model to the virtual reconstruction of the statue of Athena and, lastly, to the 3D printing of the reconstructive hypothesis. Digital models were created for many of the sculpture fragments of the Sanctuary of Apollo. Saved in a database designed to store data on the sculptures, these models facilitate the study of the documentation and have proved to be extremely useful for the dissemination of the finds to the public, especially in problematic situations that limit or impede access to the evidence, as was the case during the Covid-19 pandemic.

The bronze quadriga of the National Archaeological Museum of Naples, recovered in fragments at different times in the history of the Herculaneum excavations, represent a case study of interdisciplinary importance. Its reconstruction has never been attempted so far, due to the many factors of uncertainty that weigh on its knowledge. A new study, reconstruction and enhancement project was launched in 2020-2021 by the CNR-ISMed in cooperation with the National Archaeological Museum of Naples with the aim of producing new scientific knowledge and restoring the beauty and uniqueness of the recomposed original monument to public use. The interpretative problems related to the quadriga are numerous, complex and, concerning archaeology and its methods, they span from the study of sources and excavations records to in-depth knowledge of Roman statuary, casting techniques and the iconographic and stylistic characteristics of Roman bronzes. In this scenario, the application of methodologies and technologies combined with traditional investigations represents an innovative integrated method. Thanks to the digital verification of hypotheses, it can produce, on solid scientific bases, a virtual reconstruction of the investigated monument. This not only makes the costs for scientific research more sustainable, but enables restoration work to be directed by optimising time and resources.

In recent years, the exploit of 3D data use in Archaeology and the Cultural Heritage sector in general has caused an exponential multiplication of digital content that can be viewed on the web. Nevertheless, web platforms can display a concerning dualism: on one side some contents are over-represented with the same models uploaded dozens of times even inside the same platform; on the other, the inaccessibility or absence of proper 3D documentation for certain datasets limits the usefulness of the resources. As a result of substantial funding received (mostly from public institutions) and the volume of data produced by each digitization project, the final impacts on the broader scientific community remain limited. Starting from the analysis of data published about EU-funded projects by the European Union Commission on the platform CORDIS, this research approaches the delicate issue of the unsustainability of the current 3D data life cycle. The analysis of 110 selected projects revealed a disturbing pattern: even though the EU provided funds for many projects that approached in different ways 3D data diffusion or sharing, currently only 8 of them made the data accessible.