Graphic symbols i.e. graphic entities drawn as a visual unit in a written text and representing something other or something more than a word of that text are the research object of the NOTAE project, which investigates them in the documentary practice of the late Roman State and Post-Roman Kingdoms (400-800 AD). While research results from the project are stored by filling forms resulting from the analysis of ancient documents, we argue that the availability of a navigable knowledge graph can ease the work of researchers at finding non trivial implications in data. In this paper, we propose a first version of the NOTAE Knowledge Graph, and we outline future works and possible synergies.
With the gradual introduction of subscription and the increasing use of papyrus the presence of graphic symbols became widespread in late Roman legal practice as it already was for some time in other expressions of the so-called “pragmatic literacy” (defined in [10] as the “literacy of one who has to read or write in the course of transacting any kind of business”); it continued in Post-Roman kingdoms as part of the same historical process of reception of the late antique documentary practice. A graphic symbol is meant as a graphic device composed by graphic signs (included alphabetical ones) drawn as a visual unit in a written text and representing something other or something more than a word of that text. The message it carries on is to be discovered, not only by definition, because there is no intrinsic prior relationship between the message-bearing graphic entity and the information it conveys, but also in historical terms: a huge quantity of graphic symbols of different forms (in appearance strange or common, unique or recurring, original or archetypal) was drawn by a huge quantity of persons, both literates of any degree (professional scribes, bureaucrats, economic lites) and illiterates, in different typologies of documents, with different functions and expectations, in different places, different social, cultural and economic contexts. Examples of graphic symbols are shown in Fig. 1.
For illiterates, for example, both in the Greek-Latin graphic and linguistic
koin of the late Roman state and in the Post-Roman kingdoms as long as Latin
functioned as language of vertical communication, performing graphic symbols
in their own hand certainly meant a way of taking an active part in the writing
process: in other words, this phenomenon can be seen as the “other side” of the
written world and considered as a matter of literacy in a wider sense. Not only
for this reason the project NOTAE represents the first attempt to investigate
this material from a novel perspective, as indicated in [
With the term NOTAE system, we refer to the information system that let
human experts involved in NOTAE to insert, update and query the research
insights gathered during the project activities. In particular, by manual inspection
of ancient documentary sources, researchers maintain a database of documents,
physical supports (papyrus, wooden tablet, slate, parchment) containing texts
and graphic symbols, if any, together with a wide range of related information.
Part of this activity could be improved in future by the employment of automatic
machine learning and image processing methods [
One of the foreseen outcomes for the project, is to discover geographical and
historical implications of the employment of graphic symbols, and this requires
to provide researchers with advanced query and visualization functionalities. In
this paper, in particular, we introduce the first version of the NOTAE Knowledge
Graph (KG). A KG [
By building the NOTAE KG, we aim at (i) introducing a common vocabulary for researchers in the area, (ii) sharing a common understanding of how concepts are related, (iii) enabling the reuse of domain knowledge, and (iv) making domain assumptions explicit. In addition, to build a proper KG for the NOTAE project, we also propose a graphical user interface which allows researchers (i) to explore the NOTAE KG, (ii) to search and explore relations and connections between resources, (iii) to make historical-geographical implications, and (iv) to discover new facts about the research field.
In this paper, we envision a tool to be employed in the broad field of digital humanities, conveying information not only from the NOTAE project but also from the vast set of related repositories and services available on the Internet. The challenge here is to enhance all of these resources, turning them into a semantically enriched ecosystem to ease information accessibility and knowledge discovery. At the best of our knowledge, even though some of the available services allow to download data as Linked Open Data (LOD), none of them provides an integrated view of the research field nor allow to draw historical implications and explore data in a simple and intuitive manner. The availability of a tool such as the one proposed in this paper would foster collaboration among the researchers in the area, and could attract curious [9], casual, users by easing the diffusion of niche topics like those regarding ancient documentary texts on which NOTAE project focuses.
The rest of the work is organized as follows. Section 2 reviews the major information systems and tools in the area of ancient written sources and related historical data. Section 3 describes the first version of the NOTAE KG and its visual interface. Finally, Section 4 concludes the paper also highlighting future works. 2
The world of digital humanities offers several different digital services and Web applications providing information about ancient documents, historical facts and geographical entities, many of which are based and/or can be downloaded as semantically enriched datasets following the LOD principles. The ones that can potentially be of interest for the historical context considered by the NOTAE project are listed below.
Pleiades3 [
databases. It models relationship between documents from different sources using RDF triples.
MANTIS is the semantically enriched database of the American Numismatic Society5, a research institute devoted to study coins from all periods and cultures. In MANTIS, each record can be exported as JSON-LD, Turtle RDF/XML, and many others.
Open Context6 collects, among other resources, archaeological reports. The data can be exported in tabular (CSV) or Geo-JSON form.
Trismegistos [
EDH8 is a search tool for Latin epigraphic data. It provides a data dump including RDF (inscriptions including prosopography) and GeoJSON formats.
All of them have embraced the LOD philosophy as a mean to connect their data to the broader digital cultural heritage infrastructure. Unfortunately, all of these efforts are jeopardized by the lack of a common vocabulary and the lack of integration, which are among the NOTAE KG goals.
In addition to these services, a wide set of websites is available, containing
digital reproductions of ancient original documents. Many of these websites are
provided by university departments, libraries and archives, and do not provide
standard ways to access data. An approach similar to ours, with a narrower
target, has been proposed in [
The NOTAE KG is built on top of the NOTAE Database, which consists of data manually inserted by domain experts as a result of their research activity in the context of the project. As the NOTAE Database is a relational database, an ETL process is executed daily to synchronize it with the KG.
The NOTAE KG makes use of multiple vocabularies, providing a set of classes and properties to describe the given domain. Vocabularies, expressed using the RDFS and OWL standards, make data integration easier by reducing diversity in describing things. The data model is defined incorporates our own vocabulary, described below, as well as the standard vocabularies listed below.
FOAF11 provides terms for describing people and organizations, documents associated with them, and social connections between people. We used properties from FOAF to connect images of symbols to the Symbol class. geo12 provides terms for specifying geographical coordinates. We used geo to define latitude and longitude of the instances of the class “Place”.
An RDF schema is composed of properties and classes. Tables 1 and 2 list respectively classes and properties employed to define NOTAE KG.
The data contained in the knowledge graph are stored in a Blazegraph
triplestoreand made available through a SPARQL endpoint13. For what concerns the
visualization and the exploration of the NOTAE KG14 (see Fig. 2), we adapted
the tool developed in the context of the ARCA project [
Let’s now see an example of how we can deduce historical implications from the data in the knowledge graph. Starting from two terms of interest (see Fig. 2), such as the symbol with ID 218 and Aphrodito, we can find all the entities that bind the two. In this case, the “Receipt” document. The same method can be applied to any class or instance of the graph. From this example, emerge how easy it is to find all documents that contain a symbol or to visually trace (by looking at the links that appear on the whiteboard) to the fact that a searched symbol is in a searched document. 4
In this paper, we presented the NOTAE Knowledge Graph (KG) as a tool to aid the researchers involved in the project and curious users to identify historical and geographical implications. The KG can be explored through a handy graphical Web application, which is already available for testing, and through a SPARQL endpoint, which allows the integration into external services.
Nonetheless, what has been presented in this paper, only represents the first version of a tool that is intended to fully cover the outcomes of the NOTAE project, by also integrating additional resources as those introduced in Section 2.
The final aim of the platform is to represent a reference in ancient so-called “pragmatic literacy”, covering the main target of the project, i.e., the analysis and categorization of graphic symbols and related historical and sociological aspects. 9. Mecella, M., Leotta, F., Marrella, A., Palucci, F., Seri, C., Catarci, T.: Encouraging persons to visit cultural sites through mini-games. EAI Endorsed Trans. Serious Games 4(14), e3 (2018) 10. Parkes, M.B.: Scribes, scripts and readers: studies in the communication, presentation and dissemination of medieval texts. Hambledon Press London and Rio Grande, OH (1991)