210 likes | 294 Views
Semantically Coordinated E-Market. Semantic Web Term Project Prepared by Melike Şah. 27 January 2005. Outline. Introduction Implementation Technology base Ontology Representation Inference Engine Product Search Demo Comparison Conclusion References. Introduction. Semantic Web
E N D
Semantically Coordinated E-Market Semantic Web Term Project Prepared by Melike Şah 27 January 2005
Outline • Introduction • Implementation • Technology base • Ontology Representation • Inference Engine • Product Search • Demo • Comparison • Conclusion • References Semantically Coordinated E-Market – Melike Şah
Introduction • Semantic Web • Content with formal semantics; RDF/XML, RDFS, OWL • Machines understand content (intelligent agents) • More automated functions (reasoning) • Resource Description Framework (RDF) • XML-based language to describe resources • Create metadata about document • All resources described in RDF denoted via URI • RDF model called a “triple”, (subject, predicate, object) • Statement is combination of triples Semantically Coordinated E-Market – Melike Şah
Introduction (Cont.) • RDF Schema (RDFS) • Defined relationships among classes/resources • Web Ontology Language (OWL) • Complex applications may require more capabilities • Cardinalities, property restrictions • Class operations (union, intersection,…) • E-Market uses OWL to represent product info. • Inference engine to reason on data • Purpose is effective product search with • OWL • Reasoning engine Semantically Coordinated E-Market – Melike Şah
Implementation Technology base Ontology Product Search
Technology Base • Protégé – Ontology Representation • Java-based graphical tool • OWL plugin - Ontology editing • OWL classes, properties, restrictions • RDF/XML syntax • Algernon – Inference Engine • Implemented in Java, portable with Protégé • Querying with ASK statement • Inserting facts with TELL statement Semantically Coordinated E-Market – Melike Şah
Ontology Representation with Protégé Figure 1: UML diagram of E-Market Ontology Semantically Coordinated E-Market – Melike Şah
Ontology (Cont.) • Ontology of E-Market is created with OWL with RDF/XML syntax • Protégé is used for producing • classes • properties • relations • individuals Semantically Coordinated E-Market – Melike Şah
Ontology (Cont.) • Classes are type of owl:Class and Products is broken into specialized Classes <owl:Class rdf:ID="Fishes"> <rdfs:subClassOf> <owl:Class rdf:about="#Products"/> </rdfs:subClassOf> </owl:Class> Semantically Coordinated E-Market – Melike Şah
Ontology (Cont.) • Properties product_name, price, day, month and year are owl:DatatypeProperty <owl:DatatypeProperty rdf:about="#price"> <rdfs:domain rdf:resource="#Products"/> <rdfs:range rdf:resource="http://www.w3.org/2001/XMLSchema#float"/> </owl:DatatypeProperty> Semantically Coordinated E-Market – Melike Şah
Ontology (Cont.) • Relations has_relation_vegetables and has_relation_leguminous are owl:ObjectProperty <owl:ObjectProperty rdf:ID="has_relation_vegetables"> <rdfs:range rdf:resource="#Vegetables"/> <rdfs:domain rdf:resource="#Leguminosae"/> </owl:ObjectProperty> Semantically Coordinated E-Market – Melike Şah
Ontology (Cont.) • Total of 63 individuals are generated <Fresh rdf:ID="melike_Individual_44"> <product_name rdf:datatype="http://www.w3.org/2001/XMLSchema#string"> anchovy500gr</product_name> <price rdf:datatype="http://www.w3.org/2001/XMLSchema#float">5.0</price> <month rdf:datatype="http://www.w3.org/2001/XMLSchema#int">1</month> <year rdf:datatype="http://www.w3.org/2001/XMLSchema#int">2005</year> <day rdf:datatype="http://www.w3.org/2001/XMLSchema#int">18</day> </Fresh> Semantically Coordinated E-Market – Melike Şah
Product Search • Algernon inference engine used • Purpose: effective product querying with ontologies and inference engine • Same search is tested on database-end E-Market with SQL Semantically Coordinated E-Market – Melike Şah
Product Search with Algernon • “Search all Pepsi Products” with Algernon Figure 2: Algernon Query searches all Pepsi products Semantically Coordinated E-Market – Melike Şah
Product Search with SQL • Search all Pepsi products with SQL SELECT [product_name], [price] FROM products, Colas, Pepsi WHERE [Products].[P_ID]=[Colas].[C_ID] And [Colas].[C_ID]=[Pepsi].[pepsi_ID]; Figure 3: SQL Query searches all Pepsi products Semantically Coordinated E-Market – Melike Şah
Demo • Search examples Semantically Coordinated E-Market – Melike Şah
Comparison • Data have semantics; ontology • Reason about ontology • No need to know all class names and details of information • More effective search • a query could return comparative prices • a query could offer products if the searched product is not found Semantically Coordinated E-Market – Melike Şah
Conclusion • Formal semantics • Machine usable data • Increased usability of data • Reasoning on ontology to further enhanced the search Semantically Coordinated E-Market – Melike Şah
References [1] Tim Berners-Lee, “The Semantic Web Roadmap”, W3C Draft Document, Sep. 1998, http://www.w3.org/DesignIssues/Semantic.html. [2] Tim Berners-Lee, “The Semantic Web", Oct. 30, 2000, Newsweek interview, http://www.keepmedia.com/pubs/Newsweek/2000/10/30/317242. [3]Electronic Markets - The International Journal of Electronic Commerce and Business Media, web site http://www.electronicmarkets.org/. [4]OWL Web Ontology, site of W3C, http://www.w3.org/2004/OWL/. [5]OWL Web Ontology Language Guide: W3C Recommendation 10 Feb 2004. Smith, Welty, McGuinness, eds, http://www.w3.org/TR/owl-guide/. [6] OWL Web Ontology Language Overview: W3C Recommendation 10 Feb 2004. Deborah L. McGuinness and Frank van Harmelen eds, http://www.w3.org/TR/owl-features/. [7] Reasoning Engine [8] Protégé overview, web site http://protege.stanford.edu. [9] N. F. Noy, M. Sintek, S. Decker, M. Crubezy, R. W. Fergerson, & M. A. Musen. “Creating Semantic Web Contents with Protege-2000”, IEEE Intelligent Systems 16(2):60-71, 2001. Semantically Coordinated E-Market – Melike Şah
References (Cont.) [10] J. Gennari, M. A. Musen, R. W. Fergerson, W. E. Grosso, M. Crubézy, H. Eriksson, N. F. Noy, S. W. Tu, “The Evolution of Protégé: An Environment for Knowledge-Based Systems Development”, 2002, Technical Report. [11] Stanford University, http://smi.stanford.edu/. [12]Algernon Overview, web site http://algernon-j.sourceforge.net/doc/overview.html. [13] Algernon Rule Base Programming, http://algernon-j.sourceforge.net/. [14] Algernon in Java commands, http://algernon-j.sourceforge.net/doc/commands/ [15]Algernon in Java Documentation,http://algernon-j.sourceforge.net/doc/ [16] Using ASK and TELL to interact with Algernon V3, http://www.cs.utexas.edu/users/qr/algy/ask-tell.html [17] Algernon Tutorial, forward-chaining rules, TELL statement, http://algernon-j.sourceforge.net/tutorial/3c.html [18] University of Texas at Austin Computer Sciences Department, http://www.cs.utexas.edu/ [19] Algernon Abstract Machine, http://www.cs.utexas.edu/ftp/pub/AI-Lab/tech-reports/UT-AI-TR-00-284.ps.Z Semantically Coordinated E-Market – Melike Şah