Information Technology Reference
In-Depth Information
Dai-Nihon Shiryo
Dai-Nihon Shiryo Database
[date]
[article]
[people]
[titles]
[places]
[keywords]
Fig. 8.2.
An example of chronicle. The chronicle event is labelled by the
historiographers with attributes such as the date, names of people, official titles, places,
and keywords.
4 Extracting a Network of Historical Figures
Our framework enables us to extract networks of people from a specific
viewpoint. This stage first filters records out from a database by keywords,
range of date, and/or name of historical figures, to specify the viewpoint,
and then aggregates filtered records of every year. We next extract the
networks of figures for every year on the basis of co-occurring figures in
the same event.
More specifically, we define a connection between historical figures
based on a person dependency. Equation 1 defines the person dependency
from one person (
p
) to another person (
p'
) in a particular year (
y)
. The
person dependency is based on the extension of term dependency
described by Akaishi et al. [1][2].
records
(
p
,
p
'
)
y
pd
(
p
,
p
'
)
y
records
(
p
)
(1)
y
where
records
y
(
p
,
p'
) represents the number of records for events in
which person
p
and
p'
both appear in given year
y
;
and
records
y
(p)
represents the number of records in which only
person
p
appears.
Search WWH ::
Custom Search