Content Management, Metadata & Semantic Web Keynote Address Net.ObjectDAYS 2001, Erfurt, Germany, September 11, 2001 Amit Sheth CTO/SrVP, Voquette (www.voquette.com) [formerly Founder/CEO, Taalee, www.taalee.com] Director, Large Scale Distributed Information Systems Lab, University Of Georgia (lsdis.cs.uga.edu) amit@sheth.org Metadata Extraction is a patented pending technology of Taalee, Inc. Semantic Engine and WorldModel are trademarks of Taalee. Inc. Confidential HP Enterprise Content Management – sample user requirements (from a large Financial Svcs Company) “If a new bond comes into inventory, then we should get a message, an alert...and be able to refine to say that I only have California, Oregon and Washington clients...." “In the month of July, I received 95 e-mails from my subscriptions. These e-mails included 61 that had 143 attachments that had 67 more attachments. In total therefore, I received almost 400 documents including 5 different types (HTML,PDF, Word, Rich Media, …). Even with this volume, I had subscribed to only 10 categories in the Equities area. There are a total of 26 Equity Subscription areas and a total of 166 categories to which a user can subscribe across all Product Areas.” Professional users of a traditional Content Management Product/Solution HP 2 Enterprise Content Management – sample user requirements (from a large Financial Svcs Company) The real question is, "Which sales ideas may have significant relevance to my book of business?" For example, an earnings warning on an equity rated Hold or Lower and not owned by any of my clients may not be of high relevance to me. Ideally, a relevance analysis would: Greatly reduce the volume of Product Area Ideas sent to every FA, hopefully to perhaps 10% to 20% or less of today's volume with ideas that are potentially actionable for that FA and his/her client Result in FAs reading and evaluating the Product Area Ideas, taking appropriate actions, and generating sales because the Product Area Ideas would be relevant Result in customer satisfaction because clients would understand FAs are paying attention to their needs and developing focused ideas Professional users of a traditional Content Management Product/Solution HP 3 Enterprise Content Management – sample product requirements (from a large Financial Svcs Company) “Content generation is a more complex and probably costly problem to solve ... we reportedly create about 9 million messages a month for field delivery. On average, this would mean 1,000 messages per month per ‘big user’ or perhaps only 500 to 600 per ‘little user’.…I strongly believe an analysis is in order of the nature and necessity of generated content , the establishment of content generation standards, the movement towards development and implementation of a relevance engine, … “ Director (Product Management) of a large company that uses a leading Content Management Product HP 4 New Enterprise Content Management Challenges 1. More variety and complexity More formats (MPEG, PDF, MS Office, WM, Real, AVI, etc) More types (Docs, Images -> Audio, Video, Variety of textstructured, unstructured) More sources (internal, extranet, internet, feeds) 2. Information Overload Too much data, precious little information (Relevance) 3. Creating Value from Content How to Distribute the right content to the right people as needed? (Personalization -- book of business) Customized delivery for different consumption options (mobile/desktop, devices) Insight, Decision Making (Actionable) HP 5 New Enterprise Content Management Technical Challenges 1. Aggregation Feed handlers/Agents that understand content representation and media semantics Push-pull, Web-DB-Files, Structured-Semi-structured-Unstructured data of different types 2. Homogenization and Enhancement Enterprise-wide common view Domain model, taxonomy/classification, metadata standards Semantic Metadata– created automatically if possible 3. Semantic Applications Search, personalization, directory, alerts, etc. using metadata and semantics (semantic association and correlation), for improved relevance, intelligent personalization, customization HP 6 Semantics “meaning or relationship of meanings, or relating to meaning” (Webster) is concerned with the relationship between the linguistic symbols and their meaning or real-world objects meaning and use of data (Information System) Example: Palm -> Company, Product, Technology, Tree Name, part of location (Palm Spring, Palm Beach) Semantics, Ontologies (Domain Models), Metamodels, Metadata, Content/Data HP 7 Semantics: The Next Step in the Web’s Evolution “The Web of data (and connections) with meaning in the sense that a computer program can learn enough about what the data means to process it. . . . Imagine what computers can understand when there is a vast tangle of interconnected terms and data that can automatically be followed.” (Tim Berners-Lee, Weaving the Web, 1999) A Content Management centric definition of Semantic Web: The concept that Web-accessible content can be organized and utilized semantically, rather than though syntactic and structural methods. HP 8 Organizing Content Different and Related Objectives: Search, Browse, Summarization, Association/Relationships Indexing Clustering Classification Controlled Vocabulary, Reference Data/ Dictionary/Thesaurus Metadata Knowledge Base (Entities/Objects and Relationships) HP 9 Traditional Text Categorization Customer Training Set Statistical/AI Techniques Classify Place in a taxonomy Routing/Distribution Customer Article Feed 4715 Most traditional Content Management Products support Categorization of unstructured content.. Classification of Article 4715 Standard Metadata Feed Source: iSyndicate Posted Date: 11/20/2000 HP 10 Voquette/Taalee’s Categorization & Automatic Metadata Creation Knowledge-base & Statistical/AI Techniques Taalee Training Set & KB Classify Place in a taxonomy Catalog Metadata Automated Content Enrichment (ACE) FTE Article 4715 Metadata Standard metadata Customer Training Set & KB Semantic metadata Feed Source: iSyndicate Posted Date: 11/20/2000 Company Name: France Telecom, Equant Ticker Symbol: FTE, ENT Exchange: NYSE Topic: Company News Company Analysis Conference Calls Earnings Stock Analysis ENT Company Analysis Conference Calls Earnings Stock Analysis NYSE Member Companies Market News IPOs Classification of Article 4715 Article Feed 4715 Semantic Engine™ Precise Personalization/ Syndication/Filtering Routing/Distribution Map to another taxonomy HP 11 Technologies for Organizing Content Information Retrieval/Document Indexing TF-IDF/statistical, Clustering, LSI Statistical learning/AI: Machine learning, Bayesian, Markov Chains, Neural Network Lexical, Natural language Thesaurus, Reference data, Domain models (Ontology) Information Extractors Reasoning/Inferencing: Logic based, Knowledge-based, Rule processing and Most powerful solutions require combine several of these, addressing more of the objectives HP 12 Ontology Standardizes meaning, description, representation of involved concepts/terms/attributes Captures the semantics involved via domain characteristics, resulting in semantic metadata “Ontological Commitment” forms basis for knowledge sharing and reuse Ontology provides semantic underpinning. HP 13 An Ontology Terms/Concepts (Attributes) site latitude longitude Functional Dependencies (FDs) eventDate description Disaster Hierarchies site => latitude, longitude damage damagePhoto Natural Disaster Man-made Disaster bodyWaveMagnitude numberOfDeaths conductedBy magnitude Volcano explosiveYield NuclearTest magnitude > 0 Earthquake bodyWaveMagnitude > 0 magnitude < 10 bodyWaveMagnitude < 10 Domain Rules HP 14 Controlled Vocabularies/ Classifications/Taxonomies/Ontologies WordNet Cyc The Medical Subject Headings (MeSH): NLM's controlled vocabulary used for indexing articles, for cataloging books and other holdings, and for searching MeSH-indexed databases, including MEDLINE. MeSH terminology provides a consistent way to retrieve information that may use different terminology for the same concepts. Year 2000 MeSH includes more than 19,000 main headings, 110,000 Supplementary Concept Records (formerly Supplementary Chemical Records), and an entry vocabulary of over 300,000 terms. HP 15 Open Directory Project (ODP): Classification/Taxonomy & Directory HP 16 Example 1 – Snapshots (“Jamal Anderson”) Search for ‘Jamal Anderson’ in ‘Football’ Click on first result for Jamal Anderson View the original source HTML page. Verify that the source page contains no mention of Team name and League name. They were Taalee’s valueadditions to the metadata to facilitate easier search. View metadata. Note that Team name and League name are also included in the metadata HP 17 Example 2 – Snapshots (“Gary Sheffield”) Search for ‘Gary Sheffield’ in ‘Baseball’ Click on first result for Gary Sheffield View the original source HTML page. Verify that the source page contains no mention of Team name and League name. They were Taalee’s valueadditions to the metadata to facilitate easier search. View metadata. Note that Team name and League name are also included in the metadata HP 18 Semantic Web – Intelligent Content (supported by Taalee Semantic Engine) Intelligent Content = What You Asked for + What you need to know! Related Stock News COMPANY Competition COMPANIES in INDUSTRY with Competing PRODUCTS COMPANIES in Same or Related INDUSTRY Regulations Technology Products Important to INDUSTRY or COMPANY Industry News EPA Impacting INDUSTRY or Filed By COMPANY SEC HP 19 Semantic Application – Equity Dashboard Automatic 3rd party content integration Focused relevant content organized by topic (semantic categorization) Related news not specifically asked for (Semantic Associations) Competitive research inferred automatically Automatic Content Aggregation from multiple content providers and feeds HP 20 ASP/Enterprise hosted Internal Source 1 Research Extractor Agent 1 2 World Model Consults Knowledge Base for Cisco’s competition Internal Source 2 Extractor Agent 2 3 External feeds/Web (e.g. Reuters) Extractor Agent 3 Lucent story from external feeds picked for publishing as “semantically related” to Cisco story – passed on to Dashboard Returns result: Lucent is a competitor of Cisco Story on Cisco Semantic Engine Semantic Application 4 1 Cisco story from Source 1 passed on to add semantic associations Story on Lucent Voquette Metabase Metadata centric Content Management Architecture XCM-compliant metadata, XML or other format Third-party Content Mgmt And Syndication HP 21 Semantic Technology Features Unstructured Text Content Semi-Structured Content Structured Content Audio/Video Content with associated text (transcript, journalist notes) Create a Customized "World Model" (Taxonomy Tree with customized domain attributes) Automatically homogenize content feed tags Automatically categorize unstructured text Automatically create tags based on text Itself Create and maintain a Customized Knowledge Base for any domain Automatically enhance content tags based on information beyond text Build contextually relevant custom research applications Contextual Search (an order of magnitude better than keyword-based search) Support push or pull delivery/ingestion of content Personalization/Alerts/Notifications Real Time Indexing (stories indexed for search/personalization within a minute) Provide the user with relevant information not explicitly asked for (Semantic Associations) HP 22 Along with the evolution of metadata and semantic technologies enabling the next generation of the Web, Content Management has entered the next generation of Enhanced Content Management. Confidential HP Resources/References RDF:www.w3.org/TR/REC-rdf-syntax/ ICE: www.icestandard.org Meta Object Facility (MOF) Specification, Version 1.3, September 27, 1999: http://cgi.omg.org/cgi-bin/doc?ad/99-09-05 XML Metadata Interchange (XMI) Specification, Version 1.1, October 25, 1999: http://cgi.omg.org/cgi-bin/doc?ad/9910-02 http://cgi.omg.org/cgi-bin/doc?ad/99-10-03 DAML: www.daml.org NEWSML: newsshowcase.reuters.com PRISM: www.prismstandard.org/techdev/prismspec1.asp RIXML: www.rixml.org XCM: www.vignette.com OIL: www.ontoknowledge.org/oil SEMANTICWEB: www.semanticweb.org, business.semanticweb.org VOICEXML: www.voicexml.org MPEG7: www.darmstadt.gmd.de/mobile/MPEG7/ Taalee: www.taalee.com Applied Semantics: www.appliedsemantics.com Ontoprose: www.ontoprise.com Multimedia Data Management: Using Metadata to Integrate and Apply Digital Media, Amit Sheth & Wolfgang Klas, Eds., McGraw Hill, ISBN: 0-07057735-8, 1998. Information Brokering, Vipul Kashyap & Amit Sheth, Kluwer Academic Publishers, 2001. Voquette Semantic Technology White Paper. Mysteries of Metadata, Speaker – Amit Sheth, Workshop at Content World 2001. Infoquilt Project, LSDIS lab. http://www.taalee.com http://lsdis.cs.uga.edu/~amit