June 11-13, 2002AQUA Question-Answering System1 AQUA: AQUAINT Question Answering System Project Progress Report SAIC, San Diego KSL, Stanford.

Slides:



Advertisements
Similar presentations
Ontology-Based Computing Kenneth Baclawski Northeastern University and Jarg.
Advertisements

International Technology Alliance In Network & Information Sciences International Technology Alliance In Network & Information Sciences Paul Smart, Ali.
For Friday No reading Homework –Chapter 23, exercises 1, 13, 14, 19 –Not as bad as it sounds –Do them IN ORDER – do not read ahead here.
K S L W i n e A g e n t : Testbed Application for Semantic Web Technologies Deborah McGuinness Eric Hsu Jessica Jenkins Rob McCool Sheila McIlraith Paulo.
Silberschatz, Galvin and Gagne ©2009 Operating System Concepts – 8 th Edition, Chapter 10: File-System Interface.
Semantic Web Tools for Authoring and Using Analysis Results Richard Fikes Robert McCool Deborah McGuinness Sheila McIlraith Jessica Jenkins Knowledge Systems.
Application architectures
World Wide Web1 Applications World Wide Web. 2 Introduction What is hypertext model? Use of hypertext in World Wide Web (WWW) – HTML. WWW client-server.
Architecture & Data Management of XML-Based Digital Video Library System Jacky C.K. Ma Michael R. Lyu.
Annotating Documents for the Semantic Web Using Data-Extraction Ontologies Dissertation Proposal Yihong Ding.
ITCS 6010 Natural Language Understanding. Natural Language Processing What is it? Studies the problems inherent in the processing and manipulation of.
Tools for Developing and Using DAML-Based Ontologies and Documents Richard Fikes Deborah McGuinness Sheila McIlraith Jessica Jenkins Son Cao Tran Gleb.
11/8/20051 Ontology Translation on the Semantic Web D. Dou, D. McDermott, P. Qi Computer Science, Yale University Presented by Z. Chen CIS 607 SII, Week.
OntoWeb SIG 2: Ontology Language Standards Heiner Stuckenschmidt Vrije Universiteit Amsterdam With contributions from: Ian Horrocks and Frank van Harmelen.
Managing Data Resources. File Organization Terms and Concepts Bit: Smallest unit of data; binary digit (0,1) Byte: Group of bits that represents a single.
1 Adapting BPEL4WS for the Semantic Web The Bottom-Up Approach to Web Service Interoperation Daniel J. Mandell and Sheila McIlraith Presented by Axel Polleres.
Editing Description Logic Ontologies with the Protege OWL Plugin.
FHIRFarm – How to build a FHIR Server Farm (quickly)
Professional Informatics & Quality Assurance Software Lifecycle Manager „Tools that are more a help than a hindrance”
Artificial Intelligence Research Centre Program Systems Institute Russian Academy of Science Pereslavl-Zalessky Russia.
Semantic Web Technologies Lecture # 2 Faculty of Computer Science, IBA.
CASE Tools And Their Effect On Software Quality Peter Geddis – pxg07u.
XML, DITA and Content Repurposing By France Baril.
This chapter is extracted from Sommerville’s slides. Text book chapter
12 December, 2012 Katrin Heinze, Bundesbank CEN/WS XBRL CWA1: European Filing Rules CWA1Page 1.
Semantic Interoperability Jérôme Euzenat INRIA & LIG France Natasha Noy Stanford University USA.
AQUA: AQUAINT Question Answering System SAIC, San Diego KSL, Stanford.
AQUAINT Kickoff Meeting – December 2001 Integrating Robust Semantics, Event Detection, Information Fusion, and Summarization for Multimedia Question Answering.
Ontology Development To Support AQUA Question-Answering Richard Fikes Jessica Jenkins Bill MacCartney Rob McCool Deborah McGuinness Knowledge Systems Laboratory.
Semantic Web. Course Content
CS 346 – Chapter 8 Main memory –Addressing –Swapping –Allocation and fragmentation –Paging –Segmentation Commitment –Please finish chapter 8.
December 15, 2011 Use of Semantic Adapter in caCIS Architecture.
Author: James Allen, Nathanael Chambers, etc. By: Rex, Linger, Xiaoyi Nov. 23, 2009.
CoGenTex, Inc. Ontology-based Multimodal User Interface in MOQA AQUAINT 18-Month Workshop San Diego, California Tanya Korelsky Benoit Lavoie Ted Caldwell.
Author: William Tunstall-Pedoe Presenter: Bahareh Sarrafzadeh CS 886 Spring 2015.
Spoken dialog for e-learning supported by domain ontologies Dario Bianchi, Monica Mordonini and Agostino Poggi Dipartimento di Ingegneria dell’Informazione.
Use of Hierarchical Keywords for Easy Data Management on HUBzero HUBbub Conference 2013 September 6 th, 2013 Gaurav Nanda, Jonathan Tan, Peter Auyeung,
1 Schema Registries Steven Hughes, Lou Reich, Dan Crichton NASA 21 October 2015.
The Client/Server Database Environment Ployphan Sornsuwit KPRU Ref.
BAA - Big Mechanism using SIRA Technology Chuck Rehberg CTO at Trigent Software and Chief Scientist at Semantic Insights™
MANAGING DATA RESOURCES ~ pertemuan 7 ~ Oleh: Ir. Abdul Hayat, MTI.
Object Oriented Multi-Database Systems An Overview of Chapters 4 and 5.
Ontology-Based Computing Kenneth Baclawski Northeastern University and Jarg.
Creating Creating, Maintaining Integrating Maintaining, and IntegratingUnderstandable Knowledge Bases Richard FikesDeborah McGuinnessSheila McIlraith Jessica.
Managing Data Resources. File Organization Terms and Concepts Bit: Smallest unit of data; binary digit (0,1) Byte: Group of bits that represents a single.
For Monday Read chapter 24, sections 1-3 Homework: –Chapter 23, exercise 8.
For Monday Read chapter 26 Last Homework –Chapter 23, exercise 7.
User Profiling using Semantic Web Group members: Ashwin Somaiah Asha Stephen Charlie Sudharshan Reddy.
Of 33 lecture 1: introduction. of 33 the semantic web vision today’s web (1) web content – for human consumption (no structural information) people search.
Faculty Faculty Richard Fikes Edward Feigenbaum (Director) (Emeritus) (Director) (Emeritus) Knowledge Systems Laboratory Stanford University “In the knowledge.
For Friday Finish chapter 23 Homework –Chapter 23, exercise 15.
Metadata By N.Gopinath AP/CSE Metadata and it’s role in the lifecycle. The collection, maintenance, and deployment of metadata Metadata and tool integration.
1 USC INFORMATION SCIENCES INSTITUTE EXPECT TEMPLE: TEMPLate Extension Through Knowledge Acquisition Yolanda Gil Jim Blythe Information Sciences Institute.
REST By: Vishwanath Vineet.
Chapter – 8 Software Tools.
For Monday Read chapter 26 Homework: –Chapter 23, exercises 8 and 9.
Semantic Wiki: Automating the Read, Write, and Reporting functions Chuck Rehberg, Semantic Insights.
Versatile Information Systems, Inc International Semantic Web Conference An Application of Semantic Web Technologies to Situation.
AQUAINT Mid-Year PI Meeting – June 2002 Integrating Robust Semantics, Event Detection, Information Fusion, and Summarization for Multimedia Question Answering.
GoRelations: an Intuitive Query System for DBPedia Lushan Han and Tim Finin 15 November 2011
Semantic Web. P2 Introduction Information management facilities not keeping pace with the capacity of our information storage. –Information Overload –haphazardly.
Mechanisms for Requirements Driven Component Selection and Design Automation 최경석.
Managing Data Resources File Organization and databases for business information systems.
OKBC (Open Knowledge Base Connectivity) An API For Knowledge Servers
Architecture Components
The Client/Server Database Environment
Tools for DAML-Based Services, Query Answering, and
ONTOMERGE Ontology translations by merging ontologies Paper: Ontology Translation on the Semantic Web by Dejing Dou, Drew McDermott and Peishen Qi 2003.
Information Retrieval
Tools for DAML-Based Services, Query Answering, and
Presentation transcript:

June 11-13, 2002AQUA Question-Answering System1 AQUA: AQUAINT Question Answering System Project Progress Report SAIC, San Diego KSL, Stanford

June 11-13, 2002AQUA Question-Answering System2 Project Summary

June 11-13, 2002AQUA Question-Answering System3 Collaboration with NMSU Team This spring, NMSU has been added as an AQUAINT contractor Synergies between the NMSU Team and the SAIC Team have led to a collaborative effort for AQUAINT

June 11-13, 2002AQUA Question-Answering System4 SAIC-KSL-NMSU Collaborative System QUESTION NL Query Interlingua Query KIF Query KIF Answer Interlingua Answer NL Answer ANSWER NMSU Query Processor SAIC Interlingua  KIF Translator KSL Java Theorem Prover SAIC KIF  Interlingua Translator NMSU NL Generator

June 11-13, 2002AQUA Question-Answering System5 Key Tasks—KSL Providing knowledge tools for team Ontolingua Knowledge Server (OKS) Background ontologies and lexicons JTP deductive answer determination system Developing new methods for Answer determination from large, complex KBs Using interoperating special-purpose reasoners KB partitioning so that most reasoning occurs within partitions Using Semantic Web representation languages (e.g., DAML+OIL) Providing understandable explanations for deduced answers sourcesconversionsReasoning steps, data sources, reasoning method, data conversions, … Use-specific and user-specific customization Detecting and resolving conflicts in KBs likelyProactive background testing for both likely and provable conflicts annotateInteractive tools for helping analyst correct or annotate conflicts

June 11-13, 2002AQUA Question-Answering System6 Progress to Date—KSL Developing JTP – An answer determining reasoner Developing a suite of special-purpose reasoners E.g., for determining time-dependent answers Reasoner for QA from ontological knowledge (in DAML+OIL) Produces and caches reasoning results during KB loading Will be able to accept a series of queries, without waiting for answers Usability improvements to JTP For KB development and query-answering testing and debugging Support for rapid reloading and changing of developing KBs E.g., Checkpointing and “untell” Hierarchical presentation of reasoning explanations DQL – Standard query language and QA protocol for DAML+OIL Basic framework for client/server deductive query answering Answers generated a batch at a time Formal semantics Applicable to other representation languages Being developed jointly with the DAML language design committee Will be implemented for JTP

June 11-13, 2002AQUA Question-Answering System7 Key Tasks—SAIC Ontolingua Translator Automatic translation of information from NMSU team into KIF knowledge representations Includes dynamic semantic alignment of TMR Ontology and Ontolingua Ontology Includes inferring relations that span multiple sentences Includes flagging relations with document contexts (time, location) Pre-reasoning to pre-answer known useful questions Modular, interchangeable KB systems to determine specific types of relations that are likely to be highly relevant and useful Current system is event-based, but architecture valid for other types of representations OKS Interface Transfer, load, and store KIF representations into Ontolingua KBs Maintain DB of translated documents with info about contents of each Defines what documents are relevant to specific queries Reduces burden of KB partitioning –Each document has a separate KB file –Query TMR matched to contents of known documents to determine relevant KBs –Only KBs for relevant documents loaded into JTP to determine the answer Source Credibility Assessment Develop historical records of source reliability for sources Assessment of estimated source credibility for incoming knowledge Develop methodologies for dynamic changes to credibility ratings

June 11-13, 2002AQUA Question-Answering System8 SAIC-KSL-NMSU Collaborative System QUESTION NL Query Interlingua Query KIF Query KIF Answer Interlingua Answer NL Answer ANSWER NMSU Query Processor SAIC Interlingua  KIF Translator KSL Java Theorem Prover SAIC KIF  Interlingua Translator NMSU NL Generator

June 11-13, 2002AQUA Question-Answering System9 Converting Semantic Info to Knowledge Ontologies (and ontological philosophies) differ between Ontolingua and TMR SAIC dynamically aligns the two ontologies as part of the knowledge formation process Must be dynamic process since both TMR and Ontolingua ontologies are actively changing during system development An automated mechanism for translating information in TMR into KIF relations is needed. TMR breaks text into smallest pieces SAIC must re-unify the pieces to produce more meaningful knowledge representations Next slide shows a typical example of this problem

June 11-13, 2002AQUA Question-Answering System10 Re-Unification of Knowledge TMR includes separate references for this sentence fragment for: united-states-soldier soldier-human-adult united-states-human-adult We re-unify this into a single definition that captures that these are all the same and plural. (defobject united-states-soldier (instance-of united-states-soldier person) (has-country united-states-soldier united-states) (member-of united-states-soldier us-army-special-forces)) (defobject group-of-united-states-soldier (group group-of-united-states-soldier united-states-soldier) (cardinality group-of-united-states-soldier 36)) SAMPLE TEXT: About 36 US Special Forces troops started a month of anti-terrorism training…

June 11-13, 2002AQUA Question-Answering System11 KIF Formulations from Narratives Often, useful text sources are narratives rather than unrelated compilations of facts Story comprehension must extend across sentences to the entire text of the narrative SAIC’s experience developing knowledge representations from narratives in HPKB and RKF programs provides unique and powerful capability in this area Our “event descriptor templates” provide relations needed to generate a series of event descriptors for a narrative These allow us to answer highly complex questions about complicated, real-world situations Developed and proven successful in HPKB Program Not trying to generate all possible relations from the text—only those relations that are in the event descriptors (i.e., known useful relations)

June 11-13, 2002AQUA Question-Answering System12 Why Event Templates? Using event templates dramatically decreases the work load on JTP for each query Distributes the analysis across multiple small reasoners specialized to answer specific types of questions Pre-query analysis anticipates common questions that may be asked about this document and pre-determines the answers automatically JTP’s set of multiple reasoners includes forward chaining, which may add other relations at load time rather than waiting for a specific query against the events Result should be to dramatically improve answer response time for many queries Also provides extensibility of the system because each set of relations is handled by a separate, modular KB reasoner Support for other types of inputs than narratives may replace the mini-reasoners but doesn’t change the architecture

June 11-13, 2002AQUA Question-Answering System13 Progress to Date Application of general-purpose rules that apply across a broad spectrum of instances Initial processing of event basics in place Identification of type of event, agent performing event, basic identification of object/agent acted on in event Subject-verb-object Automatic definition of specific objects from text Places, people, groups Including cardinality of groups if available in original text Location and time of event Determination (from raw text dateline, etc.) of event context location and event context time All events in this context are tagged relative to event-context- location and event-context-time

June 11-13, 2002AQUA Question-Answering System14 Automatic Knowledge Representation Status (cont.) More sophisticated event processing Interests relations General rules about actions supporting interests of agents performing them Citizenship relations Action-object events Dealing with verbs that are objects of actions in action-object-is relations Example: “pretending to do something”  action-object-is “doing something” Representation is an event for “doing-something-event” and a separate event for “pretending-the-doing-something-event” Object re-unification Recognition of previously referred to objects (from other sentences or within longer sentences) as the same object Pronoun dereferencing Multiple phrasings of the same object

June 11-13, 2002AQUA Question-Answering System15 Status of Knowledge Representation System Preliminary demo of automatic translation from TMR to KIF is up and running Our demonstration system is limited at the moment Limited ontology Limited relations Limited lexical terms The Demo System is improving quickly in its capabilities; in the meantime…

June 11-13, 2002AQUA Question-Answering System16 Demonstration System