TY - JOUR
T1 - MODELLING AND STRUCTURING NARRATIVE HISTORICAL SOURCES WITH TEMPORAL CONTEXT
AU - Ogawa, Jun
AU - Nagasaki, Kiyonori
AU - Ohmukai, Ikki
N1 - Publisher Copyright:
© Edinburgh University Press 2022.
PY - 2022
Y1 - 2022
N2 - Creating semantically well-structured data from various historical sources is a challenging task as its contents are quite complicated and less structured. Factoid Prosopography Ontology (FPO), a source-oriented data model for prosopography, tackles this problem and enables the knowledge representation of historical information. However, we highlight two issues that FPO does not sufficiently address. The first issue relates to the temporal expression. While FPO provides a way of attaching date or the date range to historical phenomena, it is based on absolute time information, which is not always given in historical sources. To overcome this, we introduce a way of expressing relative time information by describing a temporal sequence of events mentioned in sources. Second, the more important issue is how to describe the context for each entity mentioned in sources. This context information is not explicitly described in sources but plays an essential role in reading historical sources. For this, we introduced a new concept EntityInContext, which represents ‘an actor in certain context’ as a single Uniform Resource Indicator (URI), and an instance of EntityInContext is defined based on the events’ sequence described earlier. The validity of the model was clarified by creating actual data for Caesar’s Bellum Gallicum and attempting a SPARQL search.
AB - Creating semantically well-structured data from various historical sources is a challenging task as its contents are quite complicated and less structured. Factoid Prosopography Ontology (FPO), a source-oriented data model for prosopography, tackles this problem and enables the knowledge representation of historical information. However, we highlight two issues that FPO does not sufficiently address. The first issue relates to the temporal expression. While FPO provides a way of attaching date or the date range to historical phenomena, it is based on absolute time information, which is not always given in historical sources. To overcome this, we introduce a way of expressing relative time information by describing a temporal sequence of events mentioned in sources. Second, the more important issue is how to describe the context for each entity mentioned in sources. This context information is not explicitly described in sources but plays an essential role in reading historical sources. For this, we introduced a new concept EntityInContext, which represents ‘an actor in certain context’ as a single Uniform Resource Indicator (URI), and an instance of EntityInContext is defined based on the events’ sequence described earlier. The validity of the model was clarified by creating actual data for Caesar’s Bellum Gallicum and attempting a SPARQL search.
KW - ancient history
KW - knowledge graph
KW - Linked Data
KW - ontology
KW - temporal context representation
UR - http://www.scopus.com/inward/record.url?scp=85127666791&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85127666791&partnerID=8YFLogxK
U2 - 10.3366/ijhac.2022.0274
DO - 10.3366/ijhac.2022.0274
M3 - Article
AN - SCOPUS:85127666791
SN - 1753-8548
VL - 16
SP - 17
EP - 32
JO - International Journal of Humanities and Arts Computing
JF - International Journal of Humanities and Arts Computing
IS - 1
ER -