ISPRS, Vol.34, Part 2W2, “Dynamic and Multi-Dimensional GIS”, Bangkok, May 23-25, 2001 
285 
(GIS-OT) 
data’s mass, 
j Distributed 
»ortation data 
icated ocean 
Internet and 
disseminate 
ought by the 
n spoliation 
tion can be 
geographic 
ansportation 
:ion includes 
Jetail about 
which ships 
I maintained 
ansportation 
mation (sea 
id personal 
elocity, etc.), 
lintained by 
nation was 
ler marine 
artly formed 
n or ocean 
ta providers 
nent, ocean 
tment, ship 
nces mainly 
ers provide 
;y. Second, 
lands of the 
ons require 
i. 
lission, and 
ommunities 
2 a sharing 
g of ocean 
of dynamic 
by another 
a basic 
transportation data-sharing model and some distributed data 
acquisition network intermediate components to share ocean 
transportation. The basic transportation data-sharing model is the 
core model of sharing of ocean transportation information. It 
includes common features of ocean transportation data. But it is 
not a large integrated data model, it is a seed model which can be 
extracted and extended conveniently in data, data structure and 
data type (it is not include the basic data structure and data type). 
This model is an object-oriented data model. Due to the 
differences we talked above, especially the second and the third 
differences, we establish some intermediate components to 
derive data from basic data model. With the derivation, we 
achieve dynamic extension for the basic data model and form the 
new chart for application. In derivation, we adopt the standard 
rules of information generalization and some intelligent data 
extraction and mining technology, such as Artificial Intelligence, 
Rough Set, etc. to satisfy the consistence of derivation. These 
functions are carried out in Data Extension Model. 
2.2 Distributed Computing 
In ocean transportation field, transportation decisions supported 
by GIS-OTs must often be made by many stakeholder groups 
who are distributed both geographically and socially, such as 
ocean transportation monitoring and management departments, 
ship companies, ships, etc. In addition, stakeholder groups are 
often located in different tiers of the administrative hierarchy. Data 
providers and users may also be distributed. They use different 
sophisticated software to process GIS-OT data. Different 
departments have different demands to GIS-OT and use different 
combinations of computing and analysis models. We cannot 
satisfy so many differences if we still use traditional approach (All 
data and computing models located in server, users send their 
demands through client to server) to deal with transportation 
information. 
With the development of Web technology and considering the 
differences and the balance of network workload, we should 
adopt distributed computing approach. All data, computing 
models and analysis models of GIS-OT (all computing models 
and analysis models must be separated into independent 
function components, which cannot be divided more and each 
component can achieve a basic function of ocean transportation) 
are maintained by their own providers, ocean transportation 
decision makers, users, etc. and located in distributed servers. 
But these separated data and models are not messy and 
disordered. Their locations and usages are ruled by a uniform 
standard of Computing And Analysis Cooperation Service. 
Different users send their request to computing and analysis 
cooperation service, the service returns the appropriate 
combination of computing and analysis models based on user’s 
request, and organize data from distributed servers to get the 
request results. 
2.3 Acquisition And Integration Of Real-time Information 
One of the most important features of GIS-OT is real-time. The 
main research object in GIS-OT is voyaging ship; go further and 
say, is ship features of each ship and all kinds of statistical results 
of all voyaging ships. But the information are changed each 
minute, even each second, in other words, is real-time. The main 
characters of real-time data are mass and messy. And different 
department have different demand to real-time data. For example, 
ocean transportation monitoring and management department 
only need statistical historical ocean transportation data every 
day, every month or every year to analyze, make decision and 
plan, but a single ship may need real-time data of its own or other 
ships every second to make decision when it in emergency. Thus, 
there are mainly three problems to deal with real-time data. First, 
real-time data is mass data, and was collected from many land 
sensors and ship equipments that are geographic distributed. But 
the bandwidth is limited. Second, because real-time data were 
collected from different sensors, they have different formats. The 
third, different land sensors and ship equipments give different 
data accuracy, some data very accurate, but others are not. On 
the whole, how to establish quicker access data models, more 
powerful spatial data fusion techniques and dynamic routing 
algorithms to storage, retrieval, processing and analysis are 
bottleneck to deal with real-time data of GIS-OT. 
In GIS-OT, for each different kind of sensors and ship equipments, 
we rectify the equipment error of real-time source data through 
special error correction function integrated in Real-time 
Information Error Correction Service. And we also provide 
different data transform functions for different kinds of data 
acquisition sensors and equipments in Real-time Information 
Acquisition Service. Through these data transform functions, we 
transform different data formats into standard of GIS-OT and 
stored in local database as a distributed server cooperated by 
Computing And Analysis Cooperation Service. 
3. BUILD INTERNET-BASED DISTRIBUTED GIS-OT MODEL 
In general, Internet-based distributed GIS-OT mainly consists of 
three parts: Background Ocean Transportation Information 
(Including geographic information, navigation information and 
other data), Basic System of Geographic Information for Ocean 
Transportation (The basic parts of Ocean Transportation System, 
which includes the basic functional components), and 
Applications to different users. Considering the Browser/Server 
architecture, network balance of front-end system and 
background service, and the features of ocean transportation, 
distributed GIS-OT can be divided into several keys functional 
parts described as follows. The general architecture of GIS-OT is 
shown in Figure 1. 
3.1 Data Service 
This service is the basic service of GIS-OT, it includes all data 
models to storage and process ocean transportation data from all 
different sources: 
• Basic Data Sharing Model: This is the core data model; it 
represents the basic elements of ocean transportation and 
can be extended. 
• Data Extension Model: Through this model, we can extend 
Basic Data Sharing Model to satisfy the development of 
ocean transportation. 
• IHO-S57 International Transmit Standard Model. 
• Real-time Information Acquisition Service: This service 
obtains real-time data from different sensors, and 
transforms them into standard model. 
• Real-time Information Error Correction Service: This 
service rectifies the equipment errors of different real-time 
source data. 
3.2 System Service 
The system service provides background service for GIS-OT, All 
its models are independent of application, satisfy the demand of 
distributed computing environment, and manage the distributed 
system. It includes: 
• Security Management: This service satisfy multi-users 
request in distributed environment and manage access 
privilege of different user to guarantee the security of 
GIS-OT. 
• Metadata: Metadata provides more efficient approach to 
search and use distributed ocean transportation data, and 
aid to assign and balance the data and operation. 
• Event Management: This service deal with the request to 
different component. 
• Data Connection: This service provides a unitary access 
interface for heterogeneous data. 
3.3 Computing And Analysis Cooperation Service 
This service is the core service of GIS-OT. It includes all 
components about computing, analysis and statistic model of 
ocean transportation information, and achieves the common 
functions of GIS-OT. The main functional components of 
Computing and Analysis Cooperation Service are: 
• Cartography Service: Different departments provide ocean 
transportation information to system through this service.