#ISSS2017 Vienna has ended
Back To Schedule
Tuesday, July 11 • 14:30 - 15:00
3178 Context Adaptive Modeling Tool in Service Design

Log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
Nowadays, analysts and business researchers are modeling a huge number of services. To develop a service accomplishing a simple requirement is quite easy to fulfill and any currently used modeling tool seems to be fully sufficient for the purpose. But is it really true? Are we sure that we can cover (with some level of abstraction) all processes and relationships hidden in the service provision by current modeling tool? We want to elaborate on these questions further.

The situations and services we are currently trying to model become more and more complex. Complexity itself is just one part of the issue. The other one is that the services should work in different contexts. A high number of analysts create simply as many models how many contexts they are able to recognize. Thence, the question is: Are we able to mention the role of the particular service in different contexts using current modeling tools?

Let's take a simple example of a street lighting. There is a try of cities to save money by upgrading street lamps into adjustable LED equipped lamps. The idea is to dim a light when there is no pedestrian nor a driving car on the street. For this purpose, LED lamps need to be equipped with a camera that signals the LED to increase luminosity when a pedestrian or a car is identified. Moreover, once the camera is equipped, it can be used to identify free parking slots on the street, or even to help ensure public safety (passively - by identifying moving people, lighting the street more when needed/actively - by informing a police about suspicious situations).

As it can be seen, there is at first a context of an adjustable lighting to save money. Then we used same devices to identify free parking slots and lastly to ensure a public safety. Whence, we used the same devices and similar data in different contexts to create different services.

If there is an error in some service, it can influence the other services. The problem can be a dysfunction of some cameras or the recognition software resolving in bad illumination of the street, not correct recognition of free parking slots, with no active public safety. If there is just one service deployed at the beginning and the other is added later, it can influence a traffic on the network the data are going through and slow down both service or lower a service reliability.

Therefore, for context modeling, we need to use (or develop) a tool that is not only context adaptive but also accepts the context as a part of the model itself. This tool should have following features:
Enable to decompose any situation into a set of elements
Build any related model by a specific use of those elements in any context
Recognize the differences among the contexts and describe them

Use of such modeling tool for an analysis of a complex service should bring better value for all the service participants. Nevertheless, according to the last service research, services are not only more and more complex, they become more adaptive at the same time. Therefore, we need to develop a modeling tool that will be also adaptive, meaning that it can adapt the model according to the actual context without losing any knowledge, information or relationships from previous contexts. In this paper, we propose such a tool for adaptive context service modeling.

Tuesday July 11, 2017 14:30 - 15:00 CEST
2nd Floor, Room SR 125, Institut für Computertechnik,TU Wien Gußhausstraße 27-29, 1040 Wien, Austria

Attendees (1)