Information/Content Architecture…we’re all librarians now…
Information/Content Architecture:the art and science of organizing and labeling websites, intranets, online communities, and software to support findability and usability(The IA Institute, iainstitute.org)
Hard-core semantics: Web 3.0 – where we’re goingOntology-based applicationsSemantic Web/Linked Open Data
“Crowd” semantics: Web 2.0 – where we are nowTaxonomy-based structureTagged by publishers and readers
…also known as ONTOLOGIES
An ontology captures semantic information (“meaning”) by defining relationships between concepts.
The words are symbols for the concepts; the symbols can be expressed in any human language.
This is a “triple” – commonly coded in the Resource Description Framework (RDF) language defined for the Semantic Web by W3C
Ontologies are networks of concepts; hierarchy not necessary in the network. Semantic databases are “triple stores”.
végétaleanimaux
estmange
carnivore herbivore
est est
est
mange
lion antilope
Löwe antilope
ist
ist
ist
ist
Tier Pflanze
isstisst
Fleischfresser Pflanzenfresser
Why will we use ontologies?Coherent navigationFlexible entry pointsConnections (highlights related information, aids
discovery) Represents any form of information
(un-/semi-/structured)Inferencing (look for one thing, discover a related thing)Concept matching (as opposed to term matching) Integration of external contentAids disambiguationReasoning (related to machine learning or AI, not
generally expressed in simpler, standard ontologies)
SCO Ontology (Semantic Component Ontology)WSF Ontology (Web Service Framework Ontology)AGGR Ontology (Aggregation Ontology)irON Ontology (Instance Record and Object Notation Ontology)domain ontologies, to capture the concepts and relationships for the purposes of a given OSF installation, andUMBEL (optional) or other upper-level concept ontologies, used for linkages to external systems.
http://www.mkbergman.com/wp-content/themes/ai3/images/2011Posts/sco_animation.gif
Linking Open Data cloud diagram, by Richard Cyganiak and Anja Jentzsch. http://lod-cloud.net/
1 billion RDF triples 385 million extracted from the English edition of Wikipedia 665 million extracted from other language editions and links to external
datasets3.64 million “things” (concepts) have labels and abstracts in up to 97
different languages1.83 million concepts are classified in ontologies
(http://mappings.dbpedia.org/server/ontology/classes)
416,000 persons 526,000 places 106,000 music albums 60,000 films 17,500 video games 169,000 organisations 183,000 species 5,400 diseases.
DBpedia knowledge base (as-of Sept 2011)…becoming the hub of Linked Open Data…
Taxonomies are a precursor to ontologies,a way to prepare for the future
Taxonomy
FolksonomyA folksonomy is a system of classification based on collaboratively creating and managing tags to annotate and categorize content.
Also known as collaborative tagging, social classification, social indexing, and social tagging.
Folksonomies (large scale, like Flickr) produce consensus around shared vocabularies, even in the absence of a central controlled vocabulary.
CategoryThe 'category' taxonomy lets you group posts together by sorting them into various categories.
TagThe 'post_tag' taxonomy is similar to categories, but more freeform. Impromptu classification, generally displayed near posts or in the form of tag clouds.
WordPress version 3 allows fully hierarchical custom taxonomies.
LT Compass: A European project about innovationusing language technology
InstantOn-demand Translation
(“pull”)
Professional Translation/Localisation
(“push”)
trend…Consumer/End-User
Perspective
Unified Communications & Interface
Unified Access to Information
& Services
auto-translation
SpeechProcessing
Applications
ContentProcessing
Applications
MultilingualSupport
Unified Multimodal
Multiplatform Delivery
Unified Publishing &
Service Content
localisation
trend…SME/Enterprise
Perspective
trend…Multilingual Support
Language Processing ToolsCategorisers, Clustering Engines, Language Aligners, Language Analysers, Language Data, NLU/Question Answering Engines,
Semantic Technologies, Speech Processors, Terminology Extractors
Language Technology Methods & Componentsalgorithms, co-reference resolution, clustering, discourse analysis,
Hidden Markov Models, meta-data tagging, morphology segmentation, named-entity recognition, parsing, part-of-speech tagging, query
expansion, relationship extraction, signal processing, speech segmentation, stemming, taxonomies/ontologies, topic
segmentation /recognition, truecasing, word segmentation, word sense disambiguation, etc.
LT-ENABLED CONSUMER, SME & ENTERPRISE APPLICATIONS
Speech InteractionSpeech Input
Speech OutputVirtual Agents
RobotsID/Verification
Multilingual Support
Translator ToolsTranslation Memory
Advanced Leveraging Machine Translation
Content Processing
Text InputContent Creation
Search & NavigationText Mining & Analytics
Rich Media & Speech Analytics
Language Technology Applications
links to my content curation site