270 likes | 473 Views
Tool for Ontology Paraphrasing, Querying and Visualization on the Semantic Web. Project By Senthil Kumar K 200637041 III MCA (SS) . Guides Details. Project Context. Military Intelligence Reports. XML Documents. Semantic Web (Ontology) . Information Extraction. Ontology
E N D
Tool for Ontology Paraphrasing, Querying and Visualization on the Semantic Web Project By Senthil Kumar K 200637041 III MCA (SS)
Project Context Military Intelligence Reports XML Documents Semantic Web (Ontology) Information Extraction Ontology Generation Paraphrasing and Querying with Visualization Summary of Intelligence Reports
Ontology • Ontology is defined as an explicit specification of conceptualization. • It is a way of representing things and the relationship between things in such a way that a machine can understand it. • It is a mechanized way of Knowledge Representation. • Ontology is used as the Knowledge Representation format at CAIR.
Paraphrasing Context Summarizer Module
Paraphrasing Steps • Document Planning Content Selection : Specify maximum allowed distance in graph. Text Planning : Ordering of facts at various distances. • Micro Planning Micro plans are templates with numerous slots and fillers for these slots. Lexicalization : Specify single sentence for each fact. Aggregation : Multiple sentences aggregated to improve readability. • Surface Realization Grammar used to fill up missing information.
Paraphrasing Example Input Ontology Karachi XYZ_FN moved located belongs Lahore belongs Pakistan After Lexicalization XYZ_FN is moved to Karachi. Karachi belongs to Pakistan. XYZ_FN is located at Lahore. Lahore belongs to Pakistan.
Paraphrasing Example (Contd.) After Aggregation XYZ_FN is located at Lahore which belongs to Pakistan. XYZ_FN is moved to Karachi which belongs to Pakistan. Final Summary XYZ_FN is located at Lahore which belongs to Pakistan and it is moved to Karachi which belongs to Pakistan.
SPARQL: • SPARQL is an RDF query language, that stands for Simple Protocol and RDF Query Language. • SPARQL allows for a query to consist of triple patterns, conjunctions, disjunctions, and optional patterns. Example: PREFIX abc: <http://example.com/exampleOntology#> SELECT ?capital ?country WHERE { ?x abc:cityname ?capital ; abc:isCapitalOf ?y . ?y abc:countryname ?country ; abc:isInContinent abc:Africa . } SPARQL query returns all country capitals in Africa
Information Visualization • Information Visualization concentrates on the use of computer-supported tools to explore large amount of abstract data. • Compact graphical presentation for visualizing large numbers of items possibly extracted from far larger datasets. • Enables users to make discoveries, decisions or explanations about patterns (trend, cluster, gap, outlier...), groups of items, or individual items.
Knowledge Visualization • Knowledge Visualization focuses on transferring insights and creating new knowledge. • Beyond the mere transfer of facts, knowledge visualization aims to further transfer insights, experiences, perspectives and predictions by using various complementary visualizations.
Existing Work and Issues • No tool available for paraphrasing from the ontology and rendering a visual representation of the same. • Query engine support with Visual representation of the output is required. • Browser based tool required to perform the above.
Proposed Work • To develop a web application for paraphrasing and querying from an ontology with visualization support for AINN group of CAIR (DRDO) which is fully browser based.
Client Web Server JUNG,JENA Library Java Library Browsers Java Applet Class File Ontology Querying and Visualization Applets OWL Files Architecture Diagram
Block Diagram Summary of Intelligence Reports Paraphrase Visualizer Visual Graph Ontology Ontology Graph OWL Graph Construction Visualizer Query Result Query Processor
Software Requirements • JDK (Java Development Kit) 1.5 • JRE 1.5 or above Supported Web Browser • JUNG (Java’s Universal Graph/Network Library) 1.7.6 • JENA (Java framework for building Semantic Web applications) 2.5.7
What JENA can / can’t do? • Can: • It provides a programmatic environment for RDF, RDFS and OWL. • It provides SPARQL processing support on ontology. • It provides Rule based Inference Engine. • Can’t: • It cant take Ontology as an input. • It doesn’t provide support for Graph based Traversals and Querying. • It doesn’t provide support for paraphrasing from the ontology.
What JUNG can / can’t do? • Can: • It provides a very basic library for creating simple graphs and visualizing it. • It provides basic layouts support for the graph rendering. • It can only understand vertices, edges and basic graph primitives • Can’t: • It cant model the Ontology as a graph. • It cant enforce the user specific constraints followed in owl specification on the graph. • It cant Rearrange the graph based on the changes made to the ontology
Major Modules • Paraphraser • Query Processor • Visualizer
Ontology Paraphrasing, Querying and Visualizing Tool Visual Graph OWL Query Result Paraphrased Summary Level – 0 DFD
OWL Query Level – 1 DFD 1 OWL Graph Constructor User OWL Graph OWL Graph 2 Paraphraser 3 Query Processor OWL Graph OWL Graph Textual Result Natural Language Summary 4 Visualizer Visual Graph
1.1 Read the OWL File and Load the Ontology OWL 1.2 Map the ontology resources to Nodes and Edges OWL Graph Constructor(1) Ontology Model OWL Nodes and Edges 1.3 Build the graph Using the nodes and edges OWL Graph
2.1 Select Contents of Interest OWL Graph 2.2 Lexicalize Sentences and Aggregate them. Paraphraser(2) Selected Facts Formed Sentences Natural Language Summary 2.3 Apply grammar to the formed sentences. OWL Sub Graph
OWL Graph 3.1 Validate the Query and Plan the Execution Query 3.2 Execute the Query by Traversing the Graph Query Processor(3) OWL Graph and Query Textual Result OWL Sub Graph
4.1 Create Renderer and layout for the Graph OWL Graph 4.2 Set Visual attributes for the OWL Nodes and Edges Visualizer(4) Renderer Renderer 4.3 Render OWL Graph Visual Graph
1) OWL Web Ontology Language Guide:http://www.w3.org/TR/owl-guide2) SPARQL:http://www.w3.org/TR/rdf-sparql-query3) “Generating Natural Language Descriptions from OWL Ontologies” - Ion Androutsopoulos1,2 and Dimitrios Galanis14) “Speech and Language Processing”, ISBN 81-7808-594-1 - Daniel Jurafsky and James H.Martin5) “Systemic Functional Linguistics” - http://www.isfla.org/Systemics/6) “Systemic Functional Grammar”http://minerva.ling.mq.edu.au/resource/VirtuallLibrary /Publications/sfg_firststep/SFG%20intro%20New.html7) “Paraphrasing from ontology”http://www.mindswap.org/papers/nlpowl.pdf References