|
Major Research Subjects of Arikawa, M. (1999-)
Professor
Further information is available at http://csis.u-tokyo.ac.jp/~arikawa/
Human-Centered Spatial Communication Tools to Enhance Human Spatial Capability in Daily Life
People often communicate information about places in their daily lives,
and they commonly refer to maps to find their intended places. However,
map medium is not the only one medium that represents spatial information.
Verbal media, such as texts and voices and real scene media including
pictures and videos about places are also used to communicate spatial
information. The present GIS focus mainly on geocoded data, such as
latitude and longitude and map representations, and have less function
for dealing with natural language and real scene representations which
are much more often used in enormous amounts of digital content produced
by people. New style spatial information tools focusing on human-centered
media such as texts and photos are necessary to establish smooth
communication between human and computer in dealing with spatial
information. We have been exploring such new style software tools
for assisting human's activities from the viewpoint of human-centered
spatial media.
- Spatial Album Software: PhotoField (Joint study with Hideyuki Fujita):
Location-stamped photos have been recently disseminated with the explosive
spread of GPS-equipped mobile phones. The location where a photo was
taken by a GPS-equipped mobile phone is the position of the camera
equipped in the mobile phone, not the positions of objects shown in
the photo. It makes difficult for us to tell what are shown in the photos
with only camera positions. We proposed photo vectors as a useful context
pattern of photos to solve this problem. A photo vector is composed of
the starting point as the camera's position and the ending point as some
object's position shown in a photo. Introducing the photo vectors into
digital photo management software provides users with a richer environment
to handle digital photos in spatial ways. Users can find their intended
photos via the photo vector field without browsing photos themselves.
We have developed a prototype, called PhotoField, based on this idea.
It is available from [http://www.s-it.org/photofield/]. Examples of
advanced queries using PhotoField are "find a photo taken from the opposite side of the building", "find a photo showing the left side to here" and "find a photo taken from a far position".

Fig 1. "Digital archives of pictorial diagrams displaying notable
sights in Kyoto" at the exhibition "Landscape pictures of notable
sights in Kyoto, the past and present" produced at Art Research Center,
Ritsumeikan University.
- Spatial Document Management System (SDMS) (Joint study with Yasushi Asami, Takeshi Sagara and Yoh Shiraishi):
Digital documents such as text, email, html, pdf, ms-word and ms-excel
files often include place descriptions such as addresses or place names.
SDMS, which is highly interactive software developed by us, can extract
both addresses and place names from digital documents, geocode the
extracted place words into geographic coordinates, and generate pieces
of POI (Point of Interest) which can be displayed on a map. Users only
drag and drop an icon, which corresponds to a digital document file or
a folder on the window content of SDMS in the desktop of pc. Then, SDMS
processes it, generates pieces of POI and displays them on a map. SDMS
realizes a user-friendly interface to enable users to easily deal with
all digital documents as spatial data. It can also export the generated
set of pieces of POI as Shape or G-XML formatted files to be ready for
common GIS. The functions of sorting files placed in folder windows or
desktops are necessary to find and manage their intended digital files.
The sorting functions on the present desktops are mainly provided in
alphanumeric-order or time-order. In the near future, the desktop will
provide a new promising sorting function for digital files; they will
be sorted in space-order, and this is exactly the primary function
of SDMS.

Fig 2. An example of user interface of SDMS
|