In the last few years Artificial Intelligence techniques have been strongly adapted and applied by theoretical archaeology in order to explore dynamic processes. In particular, the integrated use of GIS systems and Artificial Neural Network techniques has been recently proposed as a heuristic tool for analyzing and simulating the multifactorial relations of the so-called natural and complex settlement systems, since the GIS techniques are involved in the analysis of geographical and spatial information, while the ANNs models can learn alphanumerical information by adapting themselves to different case studies (and can display hypo- thetical solutions to specific problems testing the acquired information). The application of ANNs on the Landscape Data Bases can be intended as a Dynamic Information Architecture System and allows the identification of peculiar relationships and links between settlements and their territory in a certain historical period. Moreover, it is possible to visualize the evo- lution of settlement systems observing and comparing results in a time-line.
L’articolo propone una nuova metodologia d’analisi dei dati archeologico-territoriali fondata sui Sistemi Artificiali Adattivi come modelli della Nuova Intelligenza Artificiale. In particolare, tale proposta me- todologica sperimenta la costruzione di un modello artificiale predittivo di tipo bottom-up sul territorio di Ebla-Chora, ed elabora ipotesi relative al posizionamento geografico dei Mausolei di Nenaš: la più rilevante e ancor sconosciuta area sepolcrale del territorio di Ebla, ricordata nel celebre ‘Rituale della Regalità’ come tappa del rito di intronizzazione e luogo santo di venerazione dei sovrani defunti. Nello specifico, i siti archeologici contemporanei collocati su un dato territorio sono stati codificati in una rete di connessioni dinamiche successivamente interrogata da una serie di algoritmi di nuova generazione già applicati, con successo, in ambito epidemiologico al fine di individuare il punto d’origine di alcune malattie infettive. Il ‘modello neurale’ così concepito è stato dunque addestrato prima al riconoscimento della localizzazione delle Tombe Reali di Ur nella Regione di Ur ed Eridu in bassa Mesopotamia, poi è stato sperimentato nella Chora di Ebla per generare ipotesi sul posizionamento geografico dei Mausolei.
Where Were the Early Syrian Kings of Ebla Buried? The Ur-Eridu Survey Neural Model as an Artificial Adaptive System for the Probabilistic Localization of the Ebla Royal è madím / Ramazzotti, Marco. - In: SCIENZE DELL'ANTICHITÀ. - ISSN 1123-5713. - STAMPA. - XIX:(2013), pp. 10-34.
Where Were the Early Syrian Kings of Ebla Buried? The Ur-Eridu Survey Neural Model as an Artificial Adaptive System for the Probabilistic Localization of the Ebla Royal è madím
RAMAZZOTTI, Marco
2013
Abstract
In the last few years Artificial Intelligence techniques have been strongly adapted and applied by theoretical archaeology in order to explore dynamic processes. In particular, the integrated use of GIS systems and Artificial Neural Network techniques has been recently proposed as a heuristic tool for analyzing and simulating the multifactorial relations of the so-called natural and complex settlement systems, since the GIS techniques are involved in the analysis of geographical and spatial information, while the ANNs models can learn alphanumerical information by adapting themselves to different case studies (and can display hypo- thetical solutions to specific problems testing the acquired information). The application of ANNs on the Landscape Data Bases can be intended as a Dynamic Information Architecture System and allows the identification of peculiar relationships and links between settlements and their territory in a certain historical period. Moreover, it is possible to visualize the evo- lution of settlement systems observing and comparing results in a time-line.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.