close

The Department of Computer Science at the University of Cyprus cordially invites you to the PhD Defense entitled:

Active Ontology: Information Dynamicity Management in a Large Scale Distributed System

Speaker: Mr. Wei Xing
Affiliation: University of Cyprus, Cyprus
Category: PhD Defense
Location: Room 148, Faculty of Pure and Applied Sciences (FST-01), 1 University Avenue, 2109 Nicosia, Cyprus (directions)
Date: Friday, March 13th, 2009
Time: 13:00 - 14:00 EET
Host: Marios Dikaiakos (mdd AT cs.ucy.ac.cy)
URL: https://www.cs.ucy.ac.cy/colloquium/presentations.php#cs.ucy.pres.2009.xing

Abstract:
Ontologies are a key technology for resolving the problem of information heterogeneity. A number of ontology based information integration systems has been developed that support information management tasks. However almost all of those systems rely on a number of assumptions about the nature of environment they are used in. One of these assumptions is that the domain is more or less static. In many realistic application scenarios for ontology based information systems, however, environments are highly dynamic. This is especially true to manage heterogeneous information in large scale distributed systems, such as Grids, cloud computing. In such environments, a number of issues are raised when applying an ontology based information integration approach for information management, because of changing information, changing ontology, and changing information sources. This dissertation studies a key issue, semantic information dynamicity management in a large scale distributed environment. In our work we proposed an Active Ontology (ActOn) information model for representing dynamic information by strengthening ontology technology with a new dynamic management element. Based on the ActOn model, we developed a lifetime control and information source selection mechanisms to ensure the accurateness and consistence of the information that represented using ActOn model. Lifetime control is important for information quality. In fact, inaccurate information is an even bigger problem than missing information, because services or consumers relying on information will produce wrong actions or wrong results without warning. The selection of a suitable information source for information retrieving is particular important as the high alternative and redundancy of the available information sources in a large scale distributed environment. The idea behind ActOn model is simple: information is represented as instantiation of classes of ActOn domain ontology. In particular, it allows the value of dynamic part of information can be interpreted as an executable query; the execution of the query is controlled by its lift-time. The ActOn model comprises two kinds of ontologies, domain ontology and information source ontology, from which one can retrieve metadata to instantiate the model. A new ontology element DObject is particularly designed to link the domain ontology and the information source ontology, which in turn enable those ActOn instances be generated (or updated) dynamically in cope with the changes in a distributed environment. The life-time control mechanism is used to manage (update) dynamic information. The information source selection mechanism is used to select a suitable information source against the query execution from multiple information source candidates during the information life cycle management. To illustrate the usefulness of our ActOn model, a design and development of an ActOnbased information service for EGEE Grid is presented. The main difference with respect to other Grid information services is that it incorporate two ActOn specific modules that are not commonly found in others: a semantic information cache with lifetime management, which provides fast access to information that has been already integrated and materialized and which is still valid, and an information source selector, which is used during the generation of the execution plan for retrieving information from the information sources and allows system to adapt to changing conditions of the EGEE Grid infrastructure and to add new information sources easily. Another important aspect of our research is the study of information quality in a large scale distributed environment. We proposed a fair systematic approach to measure information quality of different information services deployed in a large scale distributed environment. Our proposed evolution framework mainly tests three key metrics of information: information completeness, consistency, and accessibility. We conducted experiments to compare and analysis two EGEE deployed information services (BDII and RGMA) and the ActOn information service on EGEE production testbed.

Short Bio:
Wei Xing is a PhD Candidate at the Department of Computer Science, University of Cyprus.

  Other Presentations Web: https://www.cs.ucy.ac.cy/colloquium/presentations.php
  Colloquia Web: https://www.cs.ucy.ac.cy/colloquium/
  Calendar: http://testing.in.cs.ucy.ac.cy/louispap/XCS-3.0/schedule/cs.ucy.pres.2009.xing.ics