Cornell CS 502 Metadata for the Web Issues and Simple Answers CS 502 – Carl Lagoze – Cornell University

Cornell CS 502 “Metadata is data about data”

Cornell CS 502 Some untested hypotheses Metadata is useful for… –People –Machines More metadata is better (semi) automated digital libraries and simple metadata

Cornell CS 502 Some known facts Number and variety of metadata vocabularies will continue to increase The Tower of Babel is a franchise –There is not one common view of reality “The one thing I know about metadata is that it is expensive”

Cornell CS 502 Are metadata and data distinguishable? Objectivity? Intellectual property? Structure? Aboutness?

Cornell CS 502 The fiction of classification …there is no classification of the universe that is not fictional and conjectural. Jorge Luis Borges

Cornell CS 502 Lenses and Views All classification does and should provide a biased lens or view of reality Each view emphasizes certain characteristics and hides others Geospatial Rights Museum

Cornell CS 502 Reality is Complex Created by: George Castaldo Created on: 1994 Created by: Leonardo da Vinci Created on: 1506 Relationship?

Cornell CS 502 Objects are Related IFLA Entity Model

Cornell CS 502 Entities, Events, and Agents Photographe r Camera type Software Computer artist

Cornell CS 502 Haven’t we done metadata already?

Cornell CS 502 What’s wrong with this model? Expensive –Complex (even for its original goal?) –Professional intervention (assumes single community of expertise) Monolithic –One size fits all approach –Reflects its centralized system origins Bias towards physical artifacts –Fixed resources –Incomplete handling of resource evolution and other resource relationships Anglo-centric

Cornell CS 502 Web Challenge to Traditional Cataloging Scale Permanence Authenticity Organizational Context Custodial Control Variety

Cornell CS 502 Internet Commons includes Multiple Communities Scientific Data Home Pages Geo Internet Commons Library Museums Commerce Whatever...

Cornell CS 502 Realities of Web search and discovery Search systems are motivated by advertising Index coverage is unpredictable and limited Too much recall, too little precision Index spam abounds Resources (and their names) are volatile

Cornell CS 502 Metadata: Part of a Solution Structured data about data –helps to impose order on chaos –enables automated discovery/manipulation Variety across various dimensions: –specialization –decentralization –democratization

Cornell CS 502 Web Metadata Issues: Description vs. Discovery Library cataloging motivated by describing resources Fuzzy search buckets –Separate books about Sigmund Freud versus books by Sigmund Freud into different buckets –But, different types of data appropriate for different buckets: URLs, date strings, word strings, names But general, fuzzy categories may not be sufficient for describing resources

Cornell CS 502 Web Metadata Models: Drill-Down Searching Paradigm Moving along a specificity spectrum Inter-domain vs. intra-domain terms, models, query mechanisms One size doesn't fit all –Cognitive models of searching and browsing

Cornell CS 502 Metadata Takes Many Forms

Cornell CS 502 Metadata: Part of the problem cost functionality AACR2/MARC google Dublin Core

Cornell CS 502 Metadata Challenges Accommodate multiple varieties of metadata –community-specific functionality, creation, administration, access Tensions –functionality and simplicity –extensibility and interoperability –human and machine creation and use

Cornell CS 502 Interoperability has many facets Semantics –Meaning/classification/ontology Models/Structure –Entities and relationships Syntax –grammars to convey semantics and structure

Cornell CS 502 Warwick Framework: Containing Chaos Conceptual Architecture for metadata from the Warwick Metadata Workshop (DC-2) Conceptual architecture to support the specification, collection, encoding, and exchange of modular metadata Provide context for metadata efforts (including Dublin Core) –avoids the “black-hole” of comprehensive element sets –focuses interoperability issues at package level

Cornell CS 502 Metadata Container Container Package Dublin Core Package MARC record Package Indirect Reference Package Terms and Conditions URI

Cornell CS 502 Modularization Allows Distributed Management Communities of expertise (not software vendors) are responsible for: –Semantics –Registration –Administration –Access management –Authority of data –Sharing and Distribution

Cornell CS 502 Modularization Implementation Issues Data encoding Semantic interaction of overlapping sets –between semantically-related packages –between semantically distinct packages Type registry

Cornell CS 502 Dublin Core Metadata Initiative A simple set of properties to support resource discovery on the web (fuzzy search buckets)? A cross-domain switchboard for interoperable metadata? An extensible ontology for resource desciption?

Cornell CS 502 The fifteen Dublin Core Elements

Cornell CS 502 A Pidgin for Digital Tourists Metadata is language Dublin Core is a small and simple language -- a pidgin -- for finding resources across domains. Speakers of different languages naturally "pidginize" to communicate –E.g., tourists using simple phrases to order beer ("zwei Bier bitte" "dva pivo" "biru o san bai"...) We are all "tourists" on the global Internet.

Cornell CS 502 A Grammar of Dublin Core r.html By design not as subtle as mother tongues, but easy to learn and extremely useful in practice Pidgins: small vocabularies (Dublin Core: fifteen special nouns and lots of optional adjectives) Simple grammars: sentences (statements) follow a simple fixed pattern...

Cornell CS 502 Example Dublin Core statements Resource has Title 'Grammar of Dublin Core'. Resource has Creator 'Tom Baker'. Resource has Subject 'Metadata'. Resource has Relation

Cornell CS 502 Resourcehasproperty DC:Creator DC:Title DC:Subject DC:Date... X implied subject implied verb one of 15 properties property value (an appropriate literal)