Probabilistic Knowledge Representation And Trust Model For Semantic Web

The World Wide Web has changed the way people communicate with each other and the way business is conducted. It lies at the heart of a revolution that is currently transforming the world toward a knowledge economy and knowledge society. However, the main obstacle to provide better support to Web users is that, at present, the meaning of Web content is not machine-accessible. The Semantic Web extends the current Web to represent Web content in a form that is more easily machine-processable and to use intelligent techniques to take advantage of these representations. The main intention of the Semantic Web that enables the Web resources to be understood by the resource usage and service mechanisms is to support high-level information extracting, knowledge fusion and knowledge-based search.The Semantic Web consists of several hierarchical layers, where the Ontology layer, in form of the OWL Web Ontology Language (recommended by the W3C), is currently the highest layer of sufficient maturity. On top of the Ontology layer, the Rules, Logic, and Proof layers of the Semantic Web will be developed next, which should offer sophisticated representation and reasoning capabilities. Following these observation, this dissertation's contributions are:â–  The author presents a novel family of representation languages that combine the expressive power of Horn rules and description logics. Following Burckert, a constrained logic scheme has been introduced with a resolution principle for Horn clauses whose variables are constrained by ALCNR description logic. Constraints can be seen as quantifier restrictions filtering out the values that any interpretation of the underlying description logic can assign to the variables of a Horn clause with such restricted quantifiers. It shows that this kind of constrained resolution is sound, complete and decidable in that a set of constrained Horn clauses is unsatisfiable over a certain description logic if and only if for each canonical interpretation of the description logic we can deduce a constrained empty clause whose constraint is satisfiable in that interpretation.â–  A hybrid system for knowledge representation, called ArtiGent, has been proposed which based on ALCNR description logic, Horn rules and probabilistic logic. It shows that the reasoning under uncertainty in the ArtiGent system is instances of a certain type of linear programming model, typically with exponentially many variables. The reasoning algorithm of the ArtiGent system has three steps: for every clash-free completion of the constraint system that is built from a certain ALCNR knowledge base, build a canonical interpretation; for each canonical interpretation, check whether we can deduce a constrained empty clausewhose constraint is satisfiable in the canonical interpretation; solve the corresponding linear optimization problems for each successful constrained resolution. We prove that the reasoning in the ArtiGent system is EXPTIME-hard.â–  Modeling trust properly and exploring techniques for establishing computational trust is a fundamental building block to realize the Semantic Web vision. A scalable probabilistic approach has been proposed for trust evaluation which combines a variety of sources of information and takes four types of costs (operational, opportunity, service charge and consultant fee) and utility into consider during the process of trust evaluation. The approach gives trust a strict probabilistic interpretation which can assist users with making the reasonable decisions in choosing the appropriate service providers according to their preferences. A formal robust analysis has been made to examine the performance of the approach.â–  Assertions of description logic and facts can be provided by relational databases and we use Semantic Database Grid to integrate database resources. Although many algorithms have been proposed for optimizing query-processing in order to minimize costs and/or response time, associated with obtaining the answer to query in distributed database systems, Semantic Database Grid query optimization problem is fundamentally different from traditional distributed query optimization. These differences are shown to be the consequences of autonomy and heterogeneity of database nodes in the database Grid. Following this observation, the author presents the design of a query optimizer for Semantic Database Grid, and proposes the query optimization algorithms DG-QOA with heuristic, dynamic, and parallel characteristics, heuristics for reducing solution space, dynamic for generating better execution sequences, and parallelism for minimizing response time.â–  Representation, acquirement and application of knowledge are three main domain of research in artificial intelligent and KR (Knowledge representation) serving as a starting point or basis of the two later. We review and discuss basic KR (logic formalisms, production system, semantic network, table, artificial neural nets, frame and etc.) and complex KR (neurules, Bayesian network, knowledge Petri net and variants), usually integrating and modifying with one or more basic. Consequently, the author proposes some trends in development of KR: hybrid or combination, methodology of Ontology, Object-Oriented and Web-Oriented. Five-levels and three dimensions of representation also have been discussed.